Four Anna Specifics – Identification and Plating 

The Four Anna lithograph value printed 1854-5 in bicolour, was more complex than other printings, and at that time at the forefront of this printing method. The individual nature of each lithograph stamp provides a wealth of diverse study, especially on an octagonal design printed in two separate stages – a separate colour for each stage. The four anna altogether has five printings, with four blue head dies recognised, in combination with two red frame dies, involving three different sheet setting/spacing types. All of which make the study of this stamp seem quite complex. However, this is not entirely the case, if remaining narrowly focused upon die identification, and watermark type with basic plating position, to reveal what printing an example might be. 

The foundation work published on the Four Anna value, establishing the above classifications was, ‘The Four Annas Lithographed Stamps of India, 1854-55 Martin D.R., Col: & E. A. Smythies: 1930 (Reprint Ed 1996)’, and is still well regarded today. Another quite important book, relating to four anna pre-issue and ‘early state’ die is, ‘The Identification of the Essays, Proofs and Reprints of the 1854 Issue of India – Spence J.A: 1975. The most recent publication and significant contribution is, ‘India’s Bi-coloured Four Annas 1854 – A Specialised Study of Third Printing – D. N. Jatia 2000. This latter work has more recently updated the overall study of the four anna, and has especially contributed much original research towards the understanding of the scarce and multi-faceted 3rd Printings. This practical guide relies heavily upon these above works that are referenced throughout. 

The original design of the value comprised of a blue head and vermillion octagonal frame printed on un-watermarked paper, and in its essay – trial printing form, were scarcer compared to the unissued lower value equivalents. The four anna essay dates to August 1854, and is primarily different to the later issued by its less finessed chignon definition, complete absence of jawline or any dotted shading on the neck, and shorter diadem band hatching lines behind ear. This initial essay head copper plate was then re-engraved at certain points, giving the sequence of the four issued head dies. Likewise, the essay frame labelled Die AI by Spence (1975), became re-engraved to Frame Die I (October 1854), and Die II (Late March 1855). The first three printings with wide spacing sheets of twelve, were bordered by blue wavy lines and rosettes in each corner.

Four anna stamps are in some key aspects less deeply researched, when compared to the two lower value mono-colour lithographs. Perhaps the two main factors for this are the relative scarcity of four anna material available to study, coupled with the complexities or impracticalities involved in differentiating the various stones on each printing, especially with head dies. Due to the broad dearth of stone reconstruction researched thus far, specific to four anna plating, then a greater significance is placed upon the die identification of each printing, watermark position and sheet spacing, in a more limited overall practical plating process. To date, only on the 3rd Printing has there been any significant progress in identifying parent stones using red frame flaws, as to do so with the blue heads is considered virtually impossible by past philatelists. 

Here is an overview of the key four anna classifications. 

First Printing Head Die I Frame Die I – 11 Stones – Wide Setting 
17,170 sheets printed - 12th to 28th October 1854 

Second Printing Head Die II Frame Die I – 11 Stones – Wide Setting 
32,830 sheets printed - 1st to 13th December 1854. 

Third Printing Head Die II - IIIA - III Frame Die I - II – 20 Stones – Wide Setting
11,580 sheets printed - 12th to 30th March 1855

Fourth Printing Head Die III Frame Die II – 10 Stones – Close Setting 
22,540 sheets printed - 29th March to 9th May 1855

Fifth Printing Head Die III Frame Die II – 8 Stones – Medium Setting
15,836 sheets printed - 4th October to 3rd November 1855

Head Die Identification [Download 22]

Primarily four anna stamps are referenced by which printing before die, because unlike the lower values they always involve combined two dies in combination. Head die identification is recommended as the first stage of the plating process. When attempting to identify the four distinct head dies (and two frame dies), it is important to recognise that each incurs a progressive state of wear. The occurrence of such progressive wear is intrinsic to the lithograph printing process, and especially effects the more finessed elements of a head die design. To process a more refined dynamic pattern recognition of the ‘Early State’ and ‘Worn State’ on each die, is very helpful. Then it is possible to comprehend the various stages of wear found in between, and even more importantly, which parts of the more standardised design on each head die will alter. As Smythies & Martin state, ‘wear is progressive there is no sharp line demarcation between the dies early and worn states’. When the stage or state of wear can be visually recognised on most four anna head dies, then a dynamic rather than rigid identification can be made. 

There are four distinct head dies – Die I (October 1854), Die II (December 1854), Die IIIA (late March 1855), Die III (late March 1855), all of which are found on wide setting/spacing printings, even before the 4th or 5th Printings with close and medium settings/printings were issued. Any alteration to a new head die was determined by the level of its wear or damage suffered during the many transfers onto various stones. 

With lithographic printing and its inherent progressive wear, it is nearly always the finer detailing of a head die design that were mostly likely to weaken first, or entirely disappear. Otherwise somewhat lighter definition not cut deeply enough on the original engraving, potentially quickened the deterioration on those features, in the consequent stages of head die usage. Head die recognition is the primary initial means to identify the correct printing, and then plate a four anna stamp. With a more refined recognition of head die, often there is little need to even glance at the frame die type. The ability to recognise at what stage a head die is at, leaves the frame die identification very much a secondary factor. In the case of a Head Die III, a stamp would be presumed 4th Printing Frame Die II, unless specific 3rd Printing frame plating flaws are found, otherwise other factors are of more significance.

Head Die I 

Key attributes: 

Early State: (1st Printing)

1. Cluster of tiny dots between hair strands around 5th Jewel.
2. Fuller chignon definition especially upper concave lines.
3. Dots on upper neck underneath unworn jawline.
4. Lines in diadem band long especially front touching upper edge.
5. Stronger and wider eyebrow.
6. Complete upper bust dotted double-line.

Late State: (1st Printing)

1. Dots between hair strands around 5th Jewel quickly disappear.
2. Weakened curved lines upper chignon forming weak comma-like motif.
3. Jawline and upper neck dots below soon wear then disappear.
4. Faded upper bust reduced to single line without dots.
5. Lines in diadem band remained unworn from original state.

The very first issued head die (October 1854) had a jawline and dots on neck added to the essay head die (August 1854), and was also given a more refined concave definition to the chignon, especially evident in the upper part. Martin and Smythies felt the progressive wear that befell Head Die I, was the most interesting aspect of the 1st Printing. Although a short printing run of 17 days meant comparatively few sheets were printed per stone, which probably averted any more-excessive wear. On Head Die I, it was thought impossible to separate with any certainty the wear of the die from that of the stone. Martin & Smythies concluded that any distinguishable variance found on Head Die I impressions, should therefore most probably be ascribed to the die itself. They surmised that the various stones encountered were wearing in the same approximate places, as demonstrated with where adjustments were made in the creation of Head Die II. 

Although all the listed Die I characteristics help reconcile stamps to this 1st Printing, some indicators are more useful than others. On any four anna stamp, a good starting place to begin head die identification is to observe the diadem band hatching. If the vertical lines at the front are still touching the upper edge or very near to it, with hatching mostly over half-length throughout the band, then it can only be Head Die I (any state) or Head Die II (early to early-middle state of wear). As the Die II upper chignon definition is entirely different even compared to a very worn state Die I, then a Head Die I 1st Printing is easily established. 

1st Printing stamps with finer shading to chignon still present, and especially tiny dots between hair strands behind 5th jewel, are a very early state indicator. Such instances are almost certainly from the first stone, or possibly no later than the second of the eleven stones considered to exist on the 1st Printing. As these finessed elements were so infrequently transferred from the die, it makes pairs in this earliest state extremely rare, to the point that Martin and Smythies felt such examples should be highly prized. Even in the very early state of wear, the jawline and upper neck shading dots below became weakened or inconsistent. Most 1st Printing stamps encountered are obviously at least slightly worn state, with inconsistent or faded jawline, although more often they are found without except short remnant of the line near to the vertical line at front of neck, and with upper neck dots reduced or mostly disappeared. 

This rapid reversion towards the jawless essay-like state without dots below, indicates an inexperience by the engraver not to etch this key feature deeply enough, as this was the first time on any issued lithograph that a head die had been given a jawline! This error was never again repeated. It seems certain that after just a few days, these useful Die I worn characteristics appeared in the majority of stones found on the 1st Printing. Therefore, an easy way to determine any 1st Printing and Head Die I, is to search for the fine dots in hair around 5th Jewel, or a slightly later but still early fading jawline, as importantly both features are not found on any other head die! In conjunction with constant longer diadem band hatching lines, and various upper chignon concave lines, this printing and head die are quite straightforward to identify. 

Head Die II

Key attributes: 

Early State: (2nd Printing)

1. Prominent comma-like motif upper chignon fine definition surrounding.
2. Strong jawline added with deeper dots below on neck.
3. Upper bust dotted double-line restored to original state.
4. Diadem band hatching lines remain firm throughout long at front. 
5. Eyebrow just beginning to fade right edge.
6. Entire outline of chignon thickened and strengthened.

Late State: (2nd Printing)

1. Comma-like motif slightly reduced less concave blank chignon surrounding. 
2. Diadem band hatching lines receding throughout half-length or less at front. 
3. Strong jawline and dots below still prominent. 
4. Eyebrow shortened fading right edge. 
5. Thickened chignon outline still prominent. 

Deterioration to the copper plate head die necessitated a recut into a new Head Die II. These 2nd Printing sheets were now inscribed in lower right margin ‘December 1854’. While some characteristics remained firm and were left untouched, certain key changes to the die were made. The master engraver no doubt realised the chignon had previously not been etched deeply enough, as a pronounced comma-like motif was now deeply entrenched into the design. As much of the original finer lines had disappeared, it remains unclear why the other already very weak chignon definition remained untouched? Perhaps at that juncture the engraver decided to redesign the chignon for longer-lasting consistency, rather than finesse?! On this new Head Die II, the jawline and dots below as well as bust outlines, were re-engraved more strongly than on Die I, and remained un-faded on all consequent dies. 

Head Die II incurred the largest quantity printed (33830 sheets) in the shortest time (13 days), with possibly as many as eleven stones made (almost one per day). Therefore, the die quickly became worn in certain places, although the newly strengthened jawline with upper neck dots below (3) remained unchanged, as did the thicker chignon outline (5), and the strengthening to the upper bust line. One notable progressive change on this head die, was the gradual wearing of the hatching lines on band of diadem. For the first time, the lines at the front were reduced to only half length in the worn state (2), with overall shortened lines before and after the overlaid hair. In a few instances, this hatching can be found very worn, faded to the extent found on a Head Die IIIA, causing some potential slight confusion with identification. There is such an example in the AIA gallery – Ex-Desai lot 624, where despite the notable reduction in length of diadem band lines, nevertheless the Head Die II chignon definition is around the middle state of wear! A more minor change was that the eyebrow faded at the right edge, and so became shortened. 
While perhaps the 2nd Printing Die II is an easier head to determine overall due to the distinctive upper chignon, this recognition is made in a more complicated context, by virtue of the fact that a small number of this head die are found on the scarce 3rd Printing. Listed as SG21a (Gibbons), these can only be determined by plating the frame flaws. 3rd Printing instances with Head Die II are usually only moderately worn and showing no need of much retouching, compared with some very worn examples found on the 2nd Printing?! Two AIA gallery examples – one single and one pair – are both with middle stage wear, as determined by the condition of diadem band hatching lines.

Does this mean progressive wear due to usage on this head die is not the only factor?’ The first stage of the 3rd Printing on the 12th March 1855, was some three months after the 2nd Printing began. Why then are the considerably later 3rd Printing not found with the very worn state Head Die II? Smythies and Martin thought perhaps there was an overlap, when Die II heads were being printed simultaneously on 2nd and 3rd Printing, including brighter frames with identifiable plating flaws. Perhaps another explanation is that an earlier transfer was made on one of the middle eleven stones, but not used at the time during the exhaustive 2nd Printing run. Being reserved, it was perhaps later brought into use for a short period, just before the transitional Head Die IIIA was produced at the end of that same month in very March 1855. 

Of all 3rd Printing types, the SG21a Head Die II is considered the most difficult to detect. Smythies and Martin felt that due to this difficulty, stamps with Head Die II would continue to go unrecognised with 3rd Printing frame flaws, ‘until the twelve positions of every stone were recorded. Since, Jatia (2000) has lessened this possibility considerably, but nevertheless it still remains a prevalent factor.

Head Die IIIA

Key attributes: 

Transitional State: (3rd Printing)

1. Comma-like motif weakened - head less bulbous.
2. Claw-like shape in upper right edge chignon (3 irregular protrusions).
3. Half star-like upper left edge chignon (3 short equidistant protrusions).
4. Incomplete comma tail weakened often into split Y-like shape.
5. Diadem band hatching lines uniformly short – very worn.
6. Two notable concave parallel lines right of weaker comma-like motif. 
7. Lower chignon lobe thickened left edge and often solid tip.

All three head die stages used in the third printing – Head Die II, Head Die IIIA, and Head Die III (Wide spacing) – present distinctive features. The Head Die IIIA in particular should be viewed as a short-lived transitional stage between Die II and Die III. Quite possibly Head Die IIIA was even used as a trial for just a few days. Martin and Smythies suggested between only one to three stones exist with this head die, although importantly the study of Jatia (2000) recorded as many as 9 of 20-22 stones, with instances of Head Die IIIA. In fact, looking at instances of head type on all 3rd Printing stones that have been recorded, strongly suggests at certain times concurrent usage of each head die, more like an overlapping transitional Head Die IIIA phase. The exact chronological sequence of each stone in progression is not understood, and so nothing is categorical.

Irrespective, the fairly minor initial upper chignon touching-up was a transitionary phase of the head die. Jatia suggests that when identifying Head Die IIIA, ‘greater stress should be laid on the clear-cut claw sign, and half ‘Y’ (see head diagram download 22), although the half star is sometimes camouflaged. The once distinctive comma-like shape is now showing new features on this die, or rather it had become in its final untouched state a progressively weakened sometimes split-ended without a bulbous head. In my view, the most important additional feature are the two prominent concave parallel vertical lines to the right of the Head Die IIIA weakened comma-like motif. These entirely new lines were probably added at this particular stage, in an intermediary transitional phase with the aim of the head to more resembling the characteristics of the original Head Die I state. This key new definition on Head Die IIIA helps avoid any potential confusion with the other two 3rd Printing head dies, especially those with uncharacteristic chignon retouches causing potential confusion. In my view, these concave parallel lines should be seen as the Primary characteristic indicator of Head Die IIIA.

One notable aspect was that the diadem band hatching was left permanently reduced, untouched from the previous Head Die II very worn state, probably for pragmatic reasons to avoid any predictable future wear. There were other slight differences made to the definition of the middle and lower lobes, such as thickened left edge and often a solid tip. Overall, these additional chignon detail changes make the transitional Head Die IIIA closer to Die III than to Die II.
Head Die III

Key attributes: 

Early State: (3rd Printing – Early 4th Printing)

1. Upper chignon finer definition added newly formed curved lines. 
2. Comma-like motif absent transformed two concave horizontal lines.
3. Diadem band hatching lines remain untouched uniformly short.
4. Jaw with dots remains constant throughout.
5. Lower chignon lobe thickened left edge often with solid tip.

Late State (Very late 4th Printing – 5th Printing): 

1. 4th crown jewel distorted mostly irregular inconsistent outline. 
2. 5th crown jewel cross weakened at centre squarer looking 4 dashes
3. Often fuller slightly erratic wispier retouched chignon definition

Although Head Die III is in itself quite easy to recognise, there is a complexity surrounding its interaction on three printings, three settings/spacing, and two frame dies. The comma-like motif of Head Die IIIA is now permanently absent, and possibly the weakened remnant split Y tail became an integral part of the Head Die III horizontally curved split definition. These newly formed concave lines of the upper chignon are the primary characteristic Head Die III indicator. Even though the chignon somewhat resembles the character of the early stages of Head Die I, it is impossible to confuse them because of the very short diadem band hatching lines now present.

On the earliest state Head Die III during the 3rd printing period, when still combined with Frame Die I, the diadem hatching looks untouched from the Head Die IIIA very worn state, as do some presumably early 4th Printing instances. Other 4th Printing examples look a little redrawn or slightly strengthened, but still with characteristic very short diadem band hatching. A change to Frame Die II (SG21c) on wide spacing was trialled for the briefest time just before the close spacing sheets were issued, between this same period. At this earlier 3rd Printing stage the crown jewels seem untouched and more worn, as most SG21 Head Die III examples show the front vertical line of the 1st jewel missing, along with general signs of wear on the other jewels. 

Certain 4th Printing examples also show this very same state of wear to the crown jewels, especially at the front. Jatia considered the jewels to generally appear in a perfect shape, and thought it a ‘positive sign’ indicating a 4th Printing. However, after a number of earlier 4th Printing Head Die III parent stones were used, when the jewels are generally worn in a certain way, quite possibly the head die was then altered again, by redrawing the crown jewels to a more perfect state, and by strengthening the permanently short vertical hatching lines of the diadem band, so they would never entirely disappear. Or in other words, the transition of the head die throughout the 3rd Printing period, continued even after the 4th Printing stamps were issued.  

The commonest use of Die III head was by far on the 4th Printing, as it was run for thirty-one days, with only minor blue retouches required throughout. This lengthier period for a lithograph printing, indicates that the die was most successful re-engraving of the copper head so far. The 4th Printing Head Die III in reality was just a progression of the state seen in the third printing. It may have had some minor adjustments made just before or even after its issue – being slightly nuanced from its 3rd Printing incarnation. However, the main alteration was the introduction of close spacing, doubling the number of stamps per sheet. 

Differentiation between the 4th Printing close spacing, and 5th Printing medium spacing both on Head Die III can be more difficult to determine, especially as there was no apparent major re-engraving of either head or frame die. The head on the 5th Printing after all is simply a later stage worn state of the same die.

Throughout the 4th Printing run the Head Die III remained remarkably consistent, as is evident by the scarcity of any retouches, when even minor instances are uncommon. However, when the 5th Printing began, the head had finally become worn, while the Die II frame held up better, having undergone considerably less transfers during its use. As the 5th Printing progressed, the head stone soon deteriorated to become very worn, necessitating many retouches including various notable examples. Therefore, retouches in general are quite a reliable 5th Printing indicator.

The 5th Printing Head Die III primary indicator, is the irregular outlined 4th jewel in crown. Martin & Smythies describe it as being of worn appearance, while Jatia states it as a distorted upper circle in an odd formation. In my view, the 4th jewel is mostly having an irregular outline, usually distorted at the sides rather than the curved top. As the 4th jewel indicator is not thought infallible, Jatia suggests using it in conjunction with the presence of the tiny red dot, found 1mm outside the left-side frame. There is also a weakening of the cross on the 5th jewel, which is in fact already evident during the later stages of the 4th Printing. More predominantly the cross becomes weakened at its centre, which now looks more square-like with four separated dashes, as the remnant of the original cross. On a later state Head Die III, other parts of the crown also become worn, including a similar wearing process on the 1st and 3rd jewels. 

Often an immediate visual indicator of Head Die III worn state, more synonymous with the 5th Printing, is a fuller often retouched with slightly erratic-looking wispier lines chignon definition. Certain other Head Die III characteristics in the worn state remain firm, such as short even hatching lines in the diadem band, and strong jaw line. Another more supplementary indicator are margins of more than 3mm wide, other than sheet marginal edges. Stamps without watermark are a fairly reliable sign of a corner sheet position of the 5th Printing, due to the accumulative medium spacing watermark alignment shift at the sheet periphery and further away from the centre. This alignment difference with 4th Printing close spacing stamps, can also be a helpful indicator between 4th and 5th Printing. 

Frame Die Identification

Even on the 1st Printing the Frame Die I wear becomes quite quickly apparent, making early state examples scarce. In the earliest state, the red dots within letters are more evident, and the middle of the outer three frame lines are consistently drawn. Too soon these central red dots especially within the A and R letters rather than D or O, begin wearing away to a miniscule size. On later printings with wide spacing these same dots often entirely disappear, becoming an obvious Frame Die I indicator. However, frame identification can occasionally be difficult due to the fact that Frame Die I and Die II characteristics can merge to some degree. This is why it is useful to comprehend what state of wear a frame die is between early and worn state. Even though emphasise is placed on the minute or missing dots in the predominantly worn state of Frame Die I, this attribute is sometimes found similarly minute on Die II, especially very late parent stones of the 5th Printing. 

On Frame Die I, the middle of the three outer frame lines remained more prominent, while the other two gradually become worn and more inconsistent. In many cases this occurrence is easily discernible, but not in all. Jatia promotes this consistent middle outer frame line as a more reliable primary indicator than the traditional method of noting the upper right diagonal hatching lines. Despite the middle outer frame line being stronger and consistent throughout, he felt this characteristic of Die I frames to be key, and should not be relegated to a secondary indicator. Jatia believed it a very reliable indicator, but importantly, only when the entire middle frame line was consistently stronger on all eight sides.

Reliance on the NE diagonal hatching lines as the primary Frame Die I indicator was thought unfounded by Jatia. He astutely observed that, ‘close examination reveals further detail of an original [hatching] characteristic’ – perhaps meaning found on Frame Die II. The deterioration of the upper right hatching was a gradual process, either caused by the transfer from the copper plate to stone, or normal wear in the course of continual printing. Jatia noted that during the 1st and 2nd, and the earlier stages of the 3rd Printing, the diagonal hatching lines are sometimes joined at their base. When studying the various stages of the frame during the 3rd Printing, he found there were incremental transfers progressing from Frame Die I into frame Die ll. The last transfer from the Die l frame from the 3rd Printing allocated the Z Stone, was by then extensively worn. Even here though the diagonal hatching lines are detached at the base, and only a few were reduced.

Only very occasionally had Jatia, ‘taken the help of diagonal (hatching) lines, but never of the dots’. Therefore, the consistent middle outer frame line, and the upper right white hatching lines should have at least equal importance as Primary Frame Die I indicators. This is especially so, when identifying frame dies on any potential more complex 3rd Printings. Martin and Smythies concluded differently, believing the two more-orthodox frame die indicators, diagonal hatching and letter centres later footnoted in Gibbons, to be sufficient. However, when both indicators are fairly equally considered, these combined characteristics should scarcely ever fail to distinguish between Frame Die I wide setting from Frame Die II close and medium settings. 

The new Frame Die II was first issued on the extremely rare very short-lived SG21c 4a wide spacing, and therefore insightful to study its original re-engraved frame state. The first transfer to stone with Frame Die II with Head Die III in the wide setting is labelled a ‘Trial Print’ by Jatia. On the first transfer to stone in close setting, there are found a few positions with a variant frame state. The masking of these very same die indicators with frame re-touches in the worn-out stages, are especially found on the 5th Printing, but occasionally the 4th Printing as well. 

Martin and Smythies describe the central red letter dots as notably strengthened throughout the 4th and 5th Printings, except in a few instances of the very last parent stones that fade considerably, somewhat similarly to typically worn Frame Die I examples. The three outer frame lines on Die II remained fairly consistent throughout with each wearing quite evenly, in great contrast to the prior frame die. This characteristic indicates the re-engraving of the frame die on all three outer lines were etched to an adequate depth. The 5th printing Frame Die II was transferred to stone much less, and so only required mostly light retouching. The inner and outer frame lines with constant usage did eventually become a little weaker at the last. 

Martin and Smythies in their 1930 monograph mention a minute red dot external to the frame, about 1 mm away on the centre-left side. This dot has been described as a guide dot, put on the stone to help space the red frame transfers, but it is so tiny and would be difficult to use as such, at a glance for convenience. If used as a guide, it would require the aid of a decent magnifying glass. Although quite a reliable 5th Printing frame indicator, apparently the tiny red dot can be found on a very late stones of the 4th Printing – so not infallible. Rather than having a specific purpose, it may have appeared accidently though Jatia, although its centralised positioning does not seem by chance?! However, this red dot is still a good secondary indicator of 5th Printing frame, especially when used in combination with other characteristic indicators, as it would only very occasionally be a 4th Printing to confuse the issue.

Trial Print Head Die III Frame Die II Wide Spacing (SG21c) 

The four anna SG21c is a special case as a trial print, with a Frame Die II still on wide setting/spacing sheet. Only an estimated 4000-5000 were printed, and they can only be detected in pairs or strips, although more recently frame flaws recorded by Jatia on 4 sheet positions makes identification of singles now more plausible. A pivotal question is why the Frame Die II was felt necessary to combine with Head Die III just before the wide spacing format was replaced by close spacing, in the very last stage of the 3rd Printing? Martin and Smythies only briefly mention this extremely rare stamp, perhaps because so little was known about it, with only a few examples as yet identified. Additionally, no reference of it is found in the official records. However, Jatia (2000) noted since then new facts have surfaced, primarily from his own research!

The 3rd Printing was a period of transition, and in the first phase stones with Head Die II were used – an obvious choice being the prior copper die was already available. Quite possibly a few earlier parent head stones were reserved in advance, or remained unused, as most 3rd Printing SG21a examples with Head Die II have a medium state of wear, rather than being the more-logical late stone very worn state. In the meantime, frame stones are presumed to have been prepared afresh. There was an adverse report from the Stamp Office concerning the differences between the original head design and the now well-worn Die II head, which induced the chief printer Thuillier to significantly re-engrave the die to more realign them. Apparently disgruntlement was specifically based around the difference of appearance in chignon definition. 

The head die re-engraving was incrementally completed in two stages of transition. Firstly, the existing Head Die II was transposed into Die IIIA, with the distinctive comma-like motif continuing to weaken, especially the tail on the left side that often split into a Y-like shape (4). The noteworthy adjustments made were various upper chignon lines added (2-3), and especially notable, two parallel concave lines just right of the now fade comma-like head (6). These two new lines should be viewed as the Primary Die characteristic indicator of Head Die IIIA. The continuously weakening diadem band hatching (5) remained untouched – probably a pragmatic decision to keep it shortened in the final transition phase, accepting the inevitable wear of the prior design. The result of these initial adjustments were then tested, combined with the frame stones already prepared. The closer alignment of the head die towards the original chignon appearance found on Head Die I had begun. This re-engraved transitional head stage Die IIIA, did not yet resemble enough the original head die. In the next transitional phase from Head Die IIIA to Die III, the copper plate was reworked again, making further changes of alignment. The fading comma-like motif was now permanently removed, replaced with a fuller less centric overall definition, now with finer newly formed concave lines. Initially it seems the newly formed Head Die III diadem band hatching was left very worn, during its 3rd Printing period of use.

The new Head Die III parent stones were made ready, giving opportunity to then proceed with the already prepared red frame stones. Head dies were apparently subject to heavier wear during printing, and so the finer details of the design needed repeated re-engraving or retouching. The frame die was sufficiently deeply cut by comparison, and without such finessed detail, enabled it to better withstand the many transfers to stone required. Therefore, not expecting any urgent need to change the red frame die, it was logical during the transitional process to leave any frame die work needed until last.

Near the end of the 3rd Printing in late March 1855, the Frame Die I had by now already undergone about a thousand transfers according to Jatia, who calculated approximately 11 stones each for 1st and 2nd Printing, and about 20 stones for the 3rd Printing, with separate transfers for the head and frame dies. Even a more robust frame die had its limits, which meant the Frame Die I finally required re-engraving. Perhaps the necessity to do so was not recognised until the listed SG21 with Head Die III (original state) had actually been run. Jatia (2000) has partly reconstructed the last 3rd Printing Frame Die I stone, plating 8 of 12 sheet positions. Labelled stone Z by him, it was useful research to better understand the final chronological sequence of frame progression between Die I to Die II. Quite possibly the reality of the frame die state of wear situation only became apparent when needing to make so many retouches, as evidenced from Z Stone plating. One significant factor noted by Jatia, is that on this final stone, most instances are Head Die II and Head Die IIIA, with only a pair with Head Die III recorded. These instances bring into question the chronological sequence in view, because of very late concurrent multiple head die usage. The Z stone had to be abandoned, and it compelled the chief printer Thuillier to undertake the creation of the rare Frame Die II with the Head Die III on a wide setting/spacing, according to Jatia. 

Perhaps it was already realised by the 15th March 1855, during the 17 day run of the entire 3rd Printing, that production was too slow. The resourceful resolution made was to change the setting from 12 to 24 stamps per sheet, and double the printing output to better meet the required demand. The necessity to re-engrave the frame to Die II before this setting/spacing change probably did not arise. The most important difference on the new Frame Die II was a deeper etching of the three outer frame lines, to make them all equally longer-lasting. This removed one of the main Frame Die I characteristics with its inconsistent weakened outer frame lines. Other alterations were, a strengthening of the upper right diagonal hatching to meet the inner white line, and a reinforcement of the centre dots, especially within the letters 'R' and the 'A'. Perhaps it was merely a practical proposition, to briefly test the new Die II frame on the wide setting format already in progress, to better appreciate how the frame adjustments actually appeared on the printed paper. It is believed this was not an unusual practice, in order to check the result of a new re-engraving. Jatia suggests this frame transition took place between the 28th and 30th March 1855, using only a single stone. Then this very small trial printing was run off with about 4000-5000 stamps, until the wide setting of 12 per sheet was immediately afterwards abandoned. 

On this extremely rare Trial Printing there are in total only three pairs and one single known, plus a single cut-to-shape copy. Two instances are sheet position 7, three position 8, two position 11, and one from position 12 (See Jatia Plate XV for plating diagrams and extant images). [Download 25] There is only a single frame stone found so far on this Trial Printing, and all sheet positions have same frame flaws. Although, the three instances from sheet position 8 clearly show different head stones, rather than the expected single stone. With just a few known examples, Jatia has managed to identify corroborating frame flaws on sheet positions 7-8 and 11-12. With the flaws recorded on positions 7 and 8, the central crease may provide some additional assistance in discovering new examples. Jatia himself found a cut-to-shape stamp from a middle row sheet position 8, using these now established invaluable plating flaws. This initial research should be very helpful in attributing further ‘Trial Printings’, including some possibly more narrow margined singles. D N Jatia, makes a strong case overall to term the SG21c wide setting as a "Trial Print".

Head and Frame Die Retouches 

On the four anna value, head die retouches should perhaps be given greater interest, as a compensation for a research deficit due to the inherent inability to plate its many head stones – quite unlike with the lower lithograph values. These often inconspicuous repairs and retouches are fascinating to detect, and detecting them is one of the charming aspects of pursuing the 1854-55 lithographic issues. However, in the modern philatelic era, it seems not so many collectors or dealers show much interest. Perhaps this is because the recognition of such retouches and classifying them, seems a more-advanced too complicated aspect of study. 

A comprehension of retouches requires a good visual recognition of what lines have been later added any given die. Often the strengthening or repair of certain elements, or even single lines, might only stand out after developing a more dynamic pattern recognition of the standardised head die characteristics. At first it can be a little difficult to gauge rare or even striking retouches, let alone minor ones. Unless armed with a good appreciation of these characteristics for each die, and where between early to worn state an example lies, then detecting any minor strokes added might be unlikely. In this respect, retouches might be determined as being uncharacteristic, depending on how well a printer has strengthened a weakened line for instance. As mentioned, when progressive state of wear is better understood, then what to look for and where to find such retouches, might suddenly become much clearer. The work of Martin and Smythies ‘The Four Annas Lithographed Stamps of India, 1854-55’, gives a very good description and overview of these various retouches found on the five four anna printings, and include 12 plates showing useful to observe retouched examples. 

Retouches of early state Head Die I are considered very rare, with only a handful in total encountered by Martin & Smythies, and apparently even in the more common worn state, retouches are still a rarity. In all, they recorded 24 different retouches of Die I Head, mostly thought more correctly to be ‘touch-ups’ (i.e. put on the stones before printing commenced). Even rarer are retouches to the frame, being it was better etched from the start, and so required less intervention. M&S illustrate a few early state examples of Die I that seem to have thickened horizontal upper chignon definition. Therefore, retouches of any kind found on the 1st Printing are very difficult to find. A notable 1st Printing sheet attribute is that left sheet positions 1, 5, 9 (column one), were stronger impressions with column two on positions 2, 6, 10, being the weakest on sheet. This sheet attribute may have had a bearing on which sheet positions were more likely to require retouches? The conclusion drawn on Die I retouched heads is that they are more likely to be a consequence of a worn die, due to excessive printing causing weak or faulty transfers onto the printing stone, rather than wear of the actual parent stones. Therefore, these should be more correctly classified as touch-ups rather than retouches, as with half Anna value.

Retouches on an early state Head Die II are also rare, as with Die I. Most are only seen around the blank parts of upper chignon, where wear had the most noticeable effect, especially as already being refined definition even in the early state. There are a very few notable instances of extensive retouching to the entire head – seen as true retouches, due to wear of stone rather than touch-ups due to wear of the Head Die II. Such examples can appear in a seemingly deeper blue ink shade, due to overall strengthening or even single retouched elements. There is a very rare entire head boldly retouched in the AIA collection that stands out compared to normal Die II heads. Minor forms of retouching are quite common to the die, such as a strengthening of tip of front jewel, with small line at diagonal running counter to design. There is also a fairly frequently encountered strengthening of tip to bust. 2nd Printing red frame retouches even in minor form are considered scarce, and slightly more notable examples are only ever seen on only about 1% of all examples, according to Martin and Smythies. 

Several extraordinary 3rd Printing retouches are found on all the three head types: Die II, Die IIIA, and Die III, and Jatia illustrates thirteen examples in colour (See Plate XX). Most are chignon retouches, or head profile. The retouched Die III heads are rightly labelled exceptional, as the die was newly created at this juncture?! The 4 of 13 illustrated retouched examples with Head Die III, each relate to crown jewels or hair. These instances probably indicate the places where the new head die was remained untouched in the initial re-engraving, and further that the same transitional process of the head continued into the early 4th Printing period! Specifically, the short diadem band hatching was later strengthened, as were the crown jewels made consistent, and probably parts of the hair, based on these 3rd Printing Head Die III very early retouches. Retouching of the chignon definition on the 3rd Printing must be seen as the most exceptional of all! While 3rd Printing retouches are rare with only a small number known, in contrast frame retouches are fairly frequent, and therefore utilised for plating. Martin and Smythies suggest this is due to faulty transfers from a worn Frame Die I, explaining why it was touched-up (or retouched) so often before the 4th Printing with Frame Die II. 

The commonest 4th Printing have virtually no major retouches, unlike any other printing. In fact, the blue head is remarkable for its scarcity of retouches, as only about 1 in 40 are found according to Smythies and Martin, and even those are nearly always only very minor. Often examples consist of a slight strengthening of profile or bust tip, or a small line in the hair of forehead, or chignon. The few instances considered major retouches on 4th Printing I feel would not perhaps be labelled such on other printings. This paucity of retouches is predictable to a degree, considering how well the Head Die III was eventually re-engraved, although its completion was perhaps only achieved during the early transfers to stone, after the 4th Printing was issued. From April until October 1855, the head die held up well with very little sign of wear or deterioration. Frame retouches on 4th Printing however are far more frequent. Martin and Smythies suggest that minor retouches are ignored, such as slight touches to inner pearls, and then the number encountered is perhaps only 30-40, with the proportion of major frame touches being very much scarcer. The principal retouched features are circle of pearls, and hatching of inner frame, while outer frame line retouches are rare. 

On the 5th Printing, about 1 in 3 stamps or possibly more have head retouches, in great contrast to the 4th Printing. This disparity reflects the fact that around October 1855 the Head Die III had finally begun to deteriorate. In fact, over half of all known four anna retouches are found on the 5th Printing alone. There is such a great variety of them, making retouches one of primary interests on this last printing. It was felt by Martin & Smythies that there were enough instances to group them thus into four definable categories: 

(i) Outline of face retouched – On at least one printing stone the nose and forehead appears to be mildly retouched on the majority of impressions, and found in conjunction with other retouches. Fairly common. 

(ii) Upper central chignon retouched – One printing stone (perhaps several stones) shows a number of extra shading lines, giving the appearance of ‘full of colour’ shading. Fairly common.  

(iii) Upper head retouched – Possibly yet another stone consisting of outline on top of head and upper hair stands redrawn with thicker cruder lines. This combination type is a part retouch part retouching-out, recorded on positions 1-4, 6-18, 21, 24. Scarce to rare. 

(iv) Whole head or large part redrawn – This type with notable extensive retouching makes the Head Die III stand out, and is considered a valuable acquisition to any collection. Various illustrations are given by Martin and Smythies (Plate 7), and recorded by them on positions 1, 2, 4, 7, 11, 17, and corner stamp. There is an excellent example on the AIA site from unrecorded position 5, showing a notable thickened head outline – the jawline especially standing out, as does the upper diadem band and boxed outline to crown jewels with fourth and especially fifth jewel obliterated! Eyebrow touching forehead outline! A combination of the above retouch types is often found on the 5th Printing. Sometimes examples are seen with other retouched features, such as the diadem band behind ear, jawline reaching the ear, spots on neck, line of bust, crown, jewels, etc. 

With these multifarious retouched stamps on the 5th Printing four anna, at least 3 of the presume 8 printing stones can be discerned. Technically, the majority of which are in reality touches-up to the head die rather than on the stone itself. On the Head Die III medium spacing, Martin & Smythies did not believe their work on the four retouched categories by any means complete. As might be expected, in the last phase of Frame Die II usage, it became considerably worn, hence the fairly frequent conspicuous retouches found.  

Die Shades

As with the lower value lithographs, on each of the five printings found on the four anna, the shade profile or spectrum encountered, can be a useful as an initial die and printing indicator. In the key publication, ‘Specialised Catalogue of Early India Lithographs 1854-1855 Dawson and Smythies 1933 2nd Ed., considerably more shade combinations were differentiated than those later listed in the Gibbons catalogues. The authors noted, ‘the shades of both blue heads and red frames are extremely variable, occurring in all possible [combinations]’. Overall, they considered no particular shade of either head or frame to be of greater value than another.  

On Head Die I issued 15th October 1854, only two shades are listed by Gibbons, blue and Indigo. On the half anna Die I, printed a little earlier between May and July 1854, there are four listed shades, blue, pale blue, deep blue, and indigo. It is probable that a very similar pigmentation and therefore shade range was in use when the four anna first run. The frequency of a Head Die I with the indigo shade encountered, i.e. SG17, is quite scarce. All other head shades found on this 1st Printing reflect the listed shades on half anna, such as dull blue, various hues of midrange blues, a few brighter blues, and deep blues. In effect all other distinct shade definition other than indigo, are dumped into the listed SG18 Blue and Pale red. Only pale blue is not really seen, but instead perhaps occasionally a pale dull blue. Four anna shade combinations tend to be quite narrow in contrast, whereas widely contrasting combinations such as deep blue & pale red, or dull blue & very deep red, are far scarcer. Dawson and Smythies (1933) listed four selective or representative shade combinations (see Other - Bibliography section), which are eight distinct shades in all, including a steel blue and vermillion combination. Steel blue is a shade akin to deeper blue or non-bright Indigo, and perhaps aligns with the latter listed label. 

Each printing also has a differing overall frame shade characteristic range, such as generally duller shades on most 1st Printings, compared to the far more varied 2nd Printing. Dawson & Smythies mention brighter shades in combination, but that is not a hue of shade I have much encountered on frames of the 1st Printing. They tend to be midrange to duller red shades, or paler reds. Their recognition of a vermillion shade, might be viewed as comparatively bright, but is probably a duller vermilion variant on this printing. On the 20th November 1854, there is a note written by Thuillier to the Government of India, wishing to obtain more vermilion (powder), as many additional sheets had been received ready for use. This pigmentation description gives some legitimacy to the shade label appearing on the 4a value 1st Printing listings. In the flesh, material with a brighter shade tendency is probably more akin to vermillion, rather than a midrange red. However, when promoting a helpful moderate shade proliferation, a considerable degree of shade standardisation is still required for pragmatic reasons. 

The 2nd Printing (December 1854) Head Die II shades have a similar overall range as the prior head die, although deeper blues and indigo are more frequently encountered. They only slightly less common than midrange blues, which tend to be duller. The blue spectrum found on this printing is still broader than the deeper shades of half anna Die II (August and November 1854). 2nd Printing Die II frame shades can vary considerably, from paler or duller shades akin to the 1st Printing, across a brighter midrange of reds, and also to deeper more intense shades, closer to the predominant frame shades found on later printings. Stamps with Head Die II and a brighter frame shade might always be a potential SG21a candidate with undetected 3rd Printing frame flaws, as brighter frame shades are far less common on the 2nd Printing. Again, Dawson and Smythies (1933) list a vermillion on this 2nd Printing, and in fact on the frame of a 3rd Printing SG21 Head Die III Frame Die I, suggesting observation of a consistent shade throughout the use of Frame Die I. There is a possibility the shade described as vermillion, may align to some degree with some frames described as ‘bright’.

Even on the short-lived scarce 3rd Printings, there is a variance of shade on the three head dies involved, ranging from midrange blues to pale blues. However, some of these although appearing at a glance to be pale, are in fact lower midrange brighter hue blues, more aligned in combination with the often seen brighter frame shades found on this printing. A true pale (non-bright blue) SG21 Pair Head Die III is found in the AIA gallery, and particularly on this head die might be seen as a precursor to the later 4th Printing SG24 listed pale blue Head Die III close spacing! Deeper blues are rarely ever seen except on perhaps very rare strongly retouched examples, giving a deeper appearance, due to applied over-inking. Rather surprisingly brighter frame shades on 3rd Printing examples are consistently encountered. They look bright when compared with frame shades mainly found on other printings. Often with these, the head dies have a brighter hue as well! Overall, this printing is never found much deeper than midrange blue and red. 

There is mostly a considerable degree of shade range overlap found between printings, as it is just that the epicentre of that spectrum shifts on each. Therefore, what shade is commonly found or infrequently found changes, sometimes with certain shades disappearing altogether. For instance, with 4th Printing frame shades, a typical shade has shifted as most have more vibrant somewhat deeper traits. The Head Die III shades encountered on the 4th Printing are better defined in the listings than on other printings, being that the range encountered is from deep to pale blues. Nevertheless, still the midrange Blue listed SG23 incorporates a broad array of sub-shades, and therefore more common when grouped together. The Frame Die II shades generally range from midrange red to deep or deeper bright reds. Often a SG22 Deep blue will frequently have a more aligned deep red frame shade, but not always. 

On the 5th Printing, Gibbons lists two shade combinations – blue and rose-red, deep blue & red, which may be seen as representative of the shade range encountered.  Quite frequently a paler slightly brighter blue is seen on this later state Head Die III. Deeper blues tend to coincide more with extensive retouches, or with slightly deeper brighter frames. Dawson and Smythies (1933) list an old rose frame shade that is later referenced by the listed rose-red. In contrast to the 4th Printing with deeper richer frame shades, many on the 5th Printing are midrange to dull with slight rose influence when compared.  

Watermark Spacing

The watermark is especially important on the four anna, in not only helping to determine the sheet position but also the printing, especially when detecting 3rd Printing candidates. There are three differently spaced sheet printings – as follows: 1st, 2nd, and 3rd Printings are wide spacing 12 stamps to sheet [Download 2], 4th Printing close spacing [Download 3], and 5th Printing medium spacing [Download 4] – both 24 stamps to sheet. Hence, three separate downloadable diagrams are required, showing the differently positioned overlay on the same watermark. Practically every wide spacing sheet has a central row horizontal fold to a greater or lesser extent. Possibly after printing runs were completed at Survey Office, they got folded in transport to post offices for distribution?! Sometimes the folds are found pressed out, to improve their condition.

Watermarks were one means of security to avoid forged copies of the official stamps. The East India company Arms watermark No: 4 sheet was deemed well-suited for the Indian lithographs, and manufactured in England was originally used for fiscal stamps. Unlike the half and one anna, on the four anna the sheet was reduced is size, and the peripheral wavy line watermark element was removed.  Only the central oval portion of the design was utilised – the sheet being a little wider in width and a little shorter in height, to incorporate the double ringed motif. On this particular value, because only a smaller sheet portion was used, overlaid with only twelve stamps now orientated in the upright position, it was easier to more accurately align the sheet when printing. 

This new partial sheet setup was quite unlike the half and one anna of ninety-six stamps that covered the full sheet in sideways orientation, with some stamps appearing without the watermark. Every four anna stamp on the sheet had a watermark on the first four printings, and only on the fifth printing (medium spacing) four corner stamps were not watermark protected. According to Jatia (2000), ‘The reason for such precise reliability of watermark can be attributed to the system adopted while laying down paper to the registration marks while printing’. While half and one anna stamps were printed in triplicate on larger sheets, on the four-anna they were split into three separate sheets. Although, still the accuracy required to consistently achieve perfect registration (or alignment) for the bi-colour lithograph stamps, proved difficult to get perfect. This is evident as various degrees of head misalignment are frequently seen, even on single sheet transfers. The need to abandon large triplicate printing sheets (triple blocks) of the four anna bi-colour is therefore obvious. 

The watermark identification task is made a little easier on the four-anna value when compared to the lower values, due to better watermark alignment overall, and every sheet position bearing a watermark, except on the medium setting. Even though slight shifts in alignment are encountered, this still makes identifying watermarks for plating purposes less problematic. Watermark orientation can cause more difficulty when establishing sheet position or sheet spacing overlay type, as four anna stamps can appear upright, inverted, reversed, and inverted & reversed. These watermark variations are in fact such a common feature that no significance of value is attached to them.

Jatia observes the relative difficulties with watermark recognition on different settings/spacing, stating ‘patience or perseverance is usually what is required’. On wide setting sheets, the central row positions 5, 6, 7, 8 can be a little problematic to determine, due to the symmetry, although correct determination of 6 and 7 can be gained by carful observation of the small inner shield top left inside cross. Perhaps this may not always be visible on a cut to shape example slightly shifted. Central positions 5 and 8 are not usually determinable by watermark alone on singles, due to the absolute symmetrical nature in design. All other sheet positions can be worked out by the non-symmetrical elements in the outer double-lined oval that would also indicate its orientation. 

On close setting watermark sheets, positions 13 and 18 can sometimes be a little difficult to determine correctly, unless the watermark is slightly shifted to show the very corner of the inscription motif, directly next to the back of the lion’s foot, in which case it can give left or right orientation (see download 3). The centre group section 8 to 11 and 14 to 17 can be more difficult to determine, unless careful note is taken of the smaller detail non-symmetrical elements, such as the small lion’s feet and shield, along with the two flags held by large lions. All sometimes tiny indicators helping to determine the correct sheet position, and watermark orientation. The careful scrutiny of whether a stamp is close or wide setting can determine if a stamp is a scarce 3rd Printing SG21, SG21c, or a far more common 4th Printing. 

On the final 5th Printing medium spacing with twenty-four stamps, sheet position identification by watermark can sometimes be difficult due to the stamp spacing on this setting. As with the 4th Printing close spacing, the double-ringed oval motif still helps in the same way on positions 2 to 5, 7, 12, 13,18, and 20 to 23 (see download 4), as do the same watermark elements found on positions 8 to 11 and 14 to 17. The four corner stamps on positions 1, 6, 19 and 24 are without the watermark if perfectly aligned to sheet, or can be just visible by the slightest favourable watermark shift. 

A useful aspect when trying to determine whether a stamp is 5th Printing, is how the alignment of the watermark falls when certain sheet positions are overlaid with medium spacing. Sometimes, this can aid identification between the two settings, depending on the sheet position and the degree of watermark shift. The further away the position is from the sheet epicentre, the more obviously the watermark will become aligned to the 5th rather than 4th Printing, due to the accumulative effect of the spacing between each stamp. The four centre stamps positions 9-10 and 15-16 would be most closely aligned between the close and medium settings, as there is barely any accumulative spacing shift. For example, on position 12 medium spacing the double rings are far more liable to show the curved lines at the very left edge of that stamp, perhaps only showing the single outer line. Whereas the same position on close spacing will probably show both curved line with a more right of stamp orientation.

Stone Identification and Plating 

With the four anna value, unfortunately there is virtually no parent stone plating research achieved, except on the red frames of the more complex 3rd Printing. Mostly it is only plausible to identify head and frame die, and watermark position, which does not constitute plating to any depth compared with the lower values. The practicalities of recording head flaws for the purpose of parent stone plating, is considered virtually impossible. Head flaws of any note are difficult to find except on the 5th Printing, and even then to find touched-up or retouched instances of both die and stone is extremely rare. Martin and Smythies said, ‘combined plating of head & frame is entirely implausible due to the enormity of permutations involved’. Only a few instances of the same head and frame flaws are known, because of the randomly combined use of red and blue stones on the same printing. It is presumed there are as many blue stones, as there are red stones. Frame flaws of any note are more frequent and easier to determine, and therefore far more practical to plate. Despite this fact, only 3rd Printing constant frame flaws have been studied to any extent.  

A reliable methodology of stone reconstruction requires an anchor point from which a definite sheet position is built upon. Then more interconnecting sheet positions need to be determined, before a stamp can be allocated with any confidence as being from the same frame die parent stone. On all five printings, pairs are very useful to lessen the logistical task, and help create a better verified interconnected stone reconstruction. The number of frame stones made on each printing, influences the difficulty factor to successfully plate a given example. Martin and Smythies believed that multiple single examples from the same sheet position were largely required to progress with a constant flaw study, especially when determining positions on a relatively large stone proliferation. For example, if 7 instances are found from the same printing on the same sheet position, all with differing frame flaws, then at least 7 frame parent stones can maybe be presumed, not accounting for sub-stones. 

After much initial effort made by earlier philatelists, there was a complete lack of success in plating head dies, although it was found plausible to deduce the approximate number of frame stones involved on each printing. In the efforts made at sheet reconstruction, a certain number of constant red frame flaws had been located in all settings, as surmised by Martin & Smythies (1930). However, on the 1st and 2nd Printing with Head Die I and II, very few instances brought success in conjoining two pairs or blocks by this reliable methodology. As head dies are known to overlap on the 3rd Printing with the same frame stones, then it might be plausible to consider that the same occurred between Die I & Die II. It was thought generally there was a lack of material extant to work out the majority of existing parent stones, and even if it were possible to gather all the pairs and blocks extant, Martin & Smythies doubted it would suffice. 

Based on various recorded instances on differing stones, together with certain positions not having retouches, initially a minimum figure of 8 Frame Die I stones were surmised on the 1st Printing. Via a different method used by Martin & Smythies, this suggested between 8 and 12 parent stones for the same printing. They then approximated the number of stones on each printing with wide setting/spacing to be about 12, with at least 36 stones noted in total. Of 180 wide setting instances including all three head dies, they found only 25 duplicated, and felt this reinforced the estimated figure of 36 stones. Hence, 12 red stones with equivalent 12 blue stones estimated on each wide spaced setting. This reconfirmed a 2nd Printing output prediction using the same method, necessitating between 10-12 stones. 

To begin with, Martin & Smythies in their monograph identified no more than 14 sets of flaws for the 3rd printing, and proceeded on that basis to recognise 14 stones. With an improved classification after new findings by Jatia were added, the number of flaws increased to 24 sets of frame flaws, for any one sheet position (see Jatia table p.32). Even if discounting some stones as having secondary flaws, still 22 stones are likely to have been used. Jatia concludes it is reasonable to assume that 20 Io 22 stones were built in all for the 3rd Printing. Between the close and medium setting/spacing of Head Die III on the 4th and 5th Printings, approximating the probable stone numbers deployed is more difficult. This is because there is greater ambiguity due to similarities of alignment on the same watermark, and the number of different red frames are also more difficult to apply to a plating process. Despite these intrinsic difficulties, Martin and Smythies identified 8 close setting/spacing stones, and 5 medium setting/spacing stones. Therefore, it was concluded that at least 8, and more probably 12 stones on close existed, and at least 5 probably 8 stones or more existed on medium spacing.

Jatia (2000) later tweaked the number of probable parent stones in his publication, but other than on the third printing, he does not elaborate upon how the number of frame stones were reached. Probably recorded from his own studies, he listed a certain number of distinct stones. 1st Printing – 11 stones, 2nd Printing – 11 stones, 3rd Printing – 20 stones, 4th Printing – 10 stones, 5th Printing – 8 stones. 

3rd Printing Plating

The most diverse 3rd Printing with smallest print run included three head dies, both frame dies, and the extremely rare SG21c as well as other rarities! The progressive order of these variations is: SG21a Head Die I Frame Die I, SG21b Head Die IIIA Frame Die I, SG21 Head Die III Frame Die I, SG21C Head Die III Frame Die II. The Head Die II on this printing can only be distinguished by red frame flaws identical to those of Head Die IIIA & III, on those same sheet positions. Head Die IIIA is a unique case, in transition from Die II to Die III. 

By 1930, the first study of 3rd Printing frame flaws by Martin and Smythies was based on 272 stamps, depicting 79 sets of confirmed flaws, which was then extended in 1951 by Smythies to 109 sets of flaws. In 2000, D N Jatia had widened the study base to about 600 stamps, with 224 identified sets of flaws, connecting to 127 stones, 65 confirmed individuals and 32 unconfirmed flaws. This later significant expansion more than doubled the initial study in every respect. Additionally, Jatia had managed to include the 3rd Printings held in the Royal collection, unlike the prior study. The enlarged 224 sets of identified frames are invaluable in resolving some of the uncertainties involved in assigning stamps to the 3rd Printing, by matching the now many more recorded frame plating flaws. The study is by no means fully comprehensive, as still a good number of missing plating components remain. 

From a total accumulation of over 4000 stamps, Martin and Smythies found 72 with Die IIIA, 200 with Die III (wide), and only 16 verified instances of Head Die II. Only 72 were actually confirmed flaws, from this 288 3rd Printing total. This considerable sample range indicated that far less SG21a Head Die II examples are recognised (rather than perhaps existing), compared to the SG21b Head Die IIIA. When calculating the relative scarcity of the three head dies when using the 3rd Printing red frame plating flaws, the main difficulty with the calculation lies with Head Die II identification, as it is more likely to be missed. In fact, it is thought many are still unrecognised. Despite these Head Die II difficulties Martin & Smythies thought that SG21a Head Die II were scarcer than SG21b Head Die IIIA, conflicting with their relative Gibbons catalogue price. I feel the SG21b is less frequently encountered than SG21a, despite the difficulty in identifying the latter. 

When reconstructing 3rd Printing parent stones, Die IIIA heads are mostly used in the initial instance of identification, hopefully later confirmed by corroborating Die III heads – according to Jatia. As the distinct Die IIIA head is detected only on 3rd printing, unlike the other two head dies, a greater number are likely to be recognised as such. On the 3rd Printing, Martin & Smythies made the calculations that there were SG21a Head Die II (2 stones) 19,000 stamps printed, SG21b Die IIIA (3 stones) 29,000 printed, SG21 Die III (9 stones) 87,000 printed, SG21c Die III Frame Die II 4000+

Modern Methodology – Plotting 3rd Printing Flaws [Download 23]

The original 1930 study of 3rd Printing frame flaws by Martin & Smythies, was constructed using a magnifying glass, and with minute findings being marked by hand on a diagram. When Smythies enlarged the scope of study of the frame flaws in 1949-51 [Download 24], he was of course unaided by modern means. Even today, philatelists often underutilise the advantages of plating with the modern advantage of digitisation. Jatia in this respect has proved well ahead of the curve, already using computer scans in the late 1990’s, to enlarge the images of each stamp for study, at a much higher level of definition than the traditional approach might offer. 

A transparent overlay of the frame design is created and placed over the original stamp, with all retouches still clearly visible underneath, and in enhanced detail. This digitised methodology made the task for Jatia far easier, as the flaws were plotted on the mask at the exact same spot and shape as found on the actual frame, making the process also very accurate. Upon blowing up each frame, he found many retouches that might once have been ignored or seemed of little consequence, now appeared more prominent. Flaws under this precise system may then be marked down without any risk of misplacement. However, digital enhancement cannot replace accumulated experience, nor provide the intuitive process required when deciding about what flaws to ignore or record. 

With any adopted methodology including this one, great caution should be taken not to allow misidentified 3rd Printing stamps into the recorded study base, otherwise the process will become corrupt. Also more caution is required in selecting the correct frame position of the sheet, when partly based on watermark that might have a different orientation. Jatia recommends to record all flaws found, as it would help in the long run. As part of the frame plating process, it is also important to note which head die types are encountered on those same sheet positions. Originally Smythies (1930), used capital letters for flaws that could definitely be designated to a frame stone, and lower case letters for confirmed flaws, which could not be assigned to any particular stone. He also allocated a separate stone, even when he came across a pair, as with the C stone positions 1-2, and R stone positions 7-8. This same method was adopted by Jatia, but he noted that, ‘Unfortunately, Smythies did not indicate the basis for connecting to a definite stone’. 

The original series of recorded Stones A, B, C, D, E, F, and R, were allocated to stones much earlier in the 1930 study, but were not put in a chronological sequence. In his subsequent enlarged 1949-51 study (see download 24), Smythies followed a slightly different system. For instance, while retaining capital letters for stones, he substituted lower case letter allocations with numerals instead. With stamps unavailable to Jatia, he adopted this later enlarged study of flaws that first appeared in the C. D. Desai auction catalogue Robson Lowe 25th May 1949, and later was reproduced in the Encyclopaedia of British Postage Stamps 1775-1950: Vol. III - The Empire in Asia - Robson Lowe: 1951 p.168. Jatia also used capital letters for stones, numerals for confirmed flaws, with a minimum of two stamp instances to keep a good integrity to his study (see download 23). Additionally, lower case letters were allotted to unconfirmed flaws, with only one recorded instance. Jatia built up groups of unconfirmed flaws by adopting this same approach. Some were eventually successfully attributed either to stones or to confirmed flaws, which explains the non-sequential lettering allocated. Underlined letters and figures indicate new additions to the stones by corresponding studies. On these highly instructive diagrams (see download 23), to recap there are three separate types of markings – Capital letters (A, B, C, etc.) for Stones, numerals (1,2,3, etc.) for Confirmed flaws, and Lower case letters (a, b, c, etc.) for Unconfirmed flaws. 

Published in 2000, the important study of Jatia had additionally identified stones, G, G1, G2, G3, H, H1, J, J1, J2, J3, J4, K, K1, K2, K3, K4, L, L1, L2, L3, M, N, N1, S, S1, S2, S3, S4, Z. (see Jatia p.45-52 - for a full descriptive list of each stone). A sequence was followed for grouping them – say position 1 and 2 were designated as G stone, 3 and 4 as H stone, 5 and 6 as J stone, and so on. Where groups had more than one set of flaws, they were earmarked as G1, G2, etc. This system has been adopted by Jatia throughout his published work. The last of the labelled sequence is the Z stone, designated as the final frame stone in the chronological sequence, just before the copper plate was re-engraved to make the Frame Die II. This is evident from the extensive flaws encountered on the Z stone, then probably considered unable to produce anymore satisfactory transfers from this existing Frame Die I.

An increase of recognised stones, identified mainly with the help of pairs and a few known larger multiples, all contribute to a sounder methodology. When plating a newly discovered stone, to create a properly anchored reconstruction, then adding any overlapping corroborated instances, much improves the integrity of the work. Far fewer vertical pairs exist than horizontal, and centre pairs on sheet positions, i.e. 2-3 or 6-7 or 10-11, are rare! This attribute of usage makes pairs that conjoin both halves of sheet, or the three rows, more difficult for plating until a scarcer bridging piece is found, possibly an occasional very scarce strip larger than a pair. Nevertheless, despite these inherent methodological difficulties, the much expanded later study by Jatia has contributed greatly to advances in our understanding of the 3rd Printing, at a deeper level than other four anna printings.  

Plating example 

Let’s say we have a single four anna stamp before us, with close cut-square margins on all four sides. From a practical viewpoint, how to proceed in its identification and potential plating. Firstly, as the example has smallish margins it could be from any printing – larger margins reduce the possible printings at a glance. Does the stamp seem pale, bright, or deeper in shade, especially the red frame? If a paler or duller red shade, it is more likely to be an earlier printing, probably wide spacing. If deeper or more intense, then maybe 2nd Printing but more likely a later 4th or 5th Printing. If the stamp is neither deep nor pale, but moderate bright red or vermillion like, as with our hypothetical example, then are we glancing at some type of scarce 3rd Printing?

A good next stage is to identify the head die type (see download 22). Is the diadem band hatching worn or shortened to any degree? If yes, then it cannot be a 1st Printing, or a 2nd Printing early state of wear. If the lines on the band are in various stages of wear, but still not short at front, then it will be either a 2nd Printing middle to worn state, or more possibly a 3rd Printing SG21a, because of the brighter non-deep frame shade. However, this stamp has worn short hatching lines from front to back, and does not have the distinct almost blank upper chignon with isolated comma-like motif at centre. Rather it has a less prominent comma-like motif with weakened left side tail, together with two curved parallel lines to right, plus some additional definition in upper parts. This makes it a probable Head Die IIIA, with the brighter non-deep frame shade indicating it as a probable 3rd Printing SG21b. 

When we now check the watermark, is it wide spacing? It depends on which part of the watermark if any shows, in relationship to the sheet spacing. Our hypothetical stamp shows one curved line indicating centre left of sheet. As the apex of the curve is central to the stamp, rather than moving in at top or bottom, then it is likely a wide setting sheet position 5 or 8 (see 3 watermark downloads 2-3-4). Then perhaps the watermark is slightly shifted east or west, hiding the second curved parallel line. It is then noticed there is a very faint horizontal crease that has likely been pressed out! Wide spacing is now confirmed; certainly the stamp is an SG21b.

Is there any need to bother trying to identify the frame die type, at this later stage in the plating process? Often it is not. Now we want to check any frame flaws for sheet positions 5 or 8, as the stamp in question may have a reversed watermark. Rather hopeful of having a rare SG21b 3rd Printing in our hands, and a little excited, next we refer to the twelve plating flaw diagrams (see Download 23) published by Jatia, which represents every sheet position found on a wide setting, clearly illustrating the known flaws on each. After studiously studying the frame under a x15 loupe, a few quite small red marks stand out from the design. Checking the diagram for sheet position 5, only one flaw similarity is partially matched. Is it enough to be allocated to the sheet position frame? Maybe not, unless other matching flaws can be found. If not, then the partially matching flaw might be considered coincidental. Frustration ensues. If the stamp has a reversed watermark, maybe matching flaws will be found on sheet position 8. Eureka, SG21b Jatia stone B, based on small red flaw top right between two outer frames, bent frame flaw above between ND below middle outer frame line. Not all noted flaws need be present, but three clearly matching flaws are a much better confirmation than two. Experience will help more often to correctly decide whether to take note or ignore similarities found. 


The early Indian four anna lithographs just like the half and one anna values has its unique attributes, when trying to plate the stamps of this fascinating multifaceted issue. The attractive bi-coloured octagonal design, coupled with the typical lithographic complexities multiplied by two, make the four anna a favourite niche collecting area for many. While the parent stones of each printing are less studied than other lithograph values, due to several pragmatic reasons such as less availability and head and frame stone identification issues, nevertheless this is partly compensated for by other facets of interest. Four anna retouches are an underappreciated area, with examples ranging from quite common to extremely rare! Then there is the elusive hunt for the almost mythical SG21c Head Die III Frame Die II wide spacing. There is no better kind of treasure hunt in Indian philately!

To properly study the four anna value, especially the complex 3rd Printing frame die flaw plating, it is a necessity to read ‘India's Bi-Coloured Four-Annas 1854 - A Specialised Study of Third Printing D N Jatia 2000’. The publication provides additional constructed tables of stones not found within this article, with confirmed and unconfirmed flaws given on each of the twelve positions, allowing an entirely new perspective. Plus, a descriptive list of sources found on each allocated stone. As well there are found many informative colour plates, illustrating the best extant material, in the context of the 3rd Printing plating study. Perhaps other four anna printings, with their various stones, might be studied in a similar way!

Introduction :: Half Anna :: One Anna :: Four Anna :: Biblio/Other

The Benns Collection
Advertise with AIA