Stamp Production

By “production” we are referring to some basic descriptive matter as linked below. Some details about the Persian and Dogri inscriptions on the stamps are also described on this page. We do not (indeed we cannot) broach technical matters about the pigments and inks. The links take us down this long page:

Papers    Embossment    Perforation    Watermarking    Plate States

Denominations of the Circulars

The currency symbols that appear at the center of the circular stamps derive from a somewhat obscure mercantile system. The images used here (of reprints, not postal issues) were chosen for their sharper impressions. The basic units of currency are related as follows:

1 rupee = 4 annas = 16 paisas = 48 pies.

We often prefer to convert everything in sight to annas, so even the 8 pies and 9 pies seen on British India stamps are occasionally rendered as 2/3 anna and 3/4 anna, respectively, the latter being equivalent to the English penny of the day.

½ anna. The group of three strokes in the center of this design really means “2 paisa.” The first stroke on the left, which is slightly curved, is a divider stroke for separating larger units to its left (here absent) from smaller units to its right, here being two vertical paisa strokes. The use of the divider stroke is analogous to the slash that separated English shillings from pence.

1 anna. Again, the curved stroke is the dividing mark that separates larger currency units from paisa units on the right, where now it is the latter that are absent. The stroke to the left of the divider is ambiguously slanted. Horizontally placed with respect to the hook it would represent 1a, while vertically-oriented it would represent ¼ rupee. Many earlier students of these stamps took these stamps to represent the latter, and the matter became something of a controversy in certain circles. The interesting philatelic history and resolution of the denominations puzzle is given in Staal pp 59-85, which is our basic reference here.

4 annas. Or, strictly speaking, ¼ rupee, and the divider stroke is absent. Older listings (but not the oldest) assumed that this is the 1 anna stamp. There are some charming philatelic pedants who insist that the 4-anna circulars are only properly called “quarter-rupee” stamps. Such folks are not usually found referring to the ½-anna as “two-paisa” stamps. So if you wish to outdo them in fussiness, you know what to do.

Inscriptions (Early Period)

The circulars and the rectangulars of the early period bear so much in common that it is advantageous to consider them in tandem with an eye on the contrasts. Dogri arcs clockwise at the top on both stamps, reading Dak Jamu on the circular and Jamu Kashmir (separated by the sun-symbol) on the rectangular. Persian arcs around the bottom on both stamps, but in opposite directions; on circulars it reads top-outward counter-clockwise and on rectangulars it reads top-inward clockwise. Word-by-word details are given down-screen.

A frozen date ‘1923’ appears on both stamps, and indeed on all stamps of the 1866-78 period. The numerals in blue are Persian and those in red are Dogri. 1923 is a year of the Hindu samvat ‘era’ that overlaps with 1866 from March 12 on. It probably marks the inauguration of the use of stamps by the State.

Red highlighting on both stamps marks ‘Jammu’ in both Dogri & Persian. One notices that the final vowel of ‘Jammu’ in the Dogri at the top is differently rendered on the two stamps. A nagari-inspired u-diacritic is used on the circular, while the formal Dogri vowel-symbol that looks like a 6 is used on the rectangular.

Blue highlighting marks ‘Kashmir’ in both Dogri & Persian. For decorative reasons the Persian on the circular is festooned with extra dots superfluous to the script, while that on the rectangular dispenses with several dots that are ordinarily wanted.

Green highlighting (on the circular only) reads dâk ~ post, mail. One might thus refer to the circulars as the “Jammu Dawks,” given that the name ‘Kashmir’ is absent in the Dogri on the circular, and dâk is absent on the rectangulars. (The earliest circular issues were however used in Kashmir as well.)

The green portion in Persian on both stamps is qalamrao ~ realm, dominion, etc. The yellow is riyâsat ~ ruling prince. The purple, top outward on the circular and top inward on the rectangular, is sarkâr ~ government. The sar is a Semitic word for ‘prince’, etc., and the kâr pertains to service, etc. So says the dictionary.

¼ anna. The yellow is Persian pâv ~ quarter. The three yellow dots would normally be written underneath the first tooth on the right to signify the p but they have been moved leftward to give space for the pair of date numerals “23” of the samvat year. The upward yellow swoop is a long-a, and the final letter that looks like a large comma is Persian vav ~ v here. The red element is Dogrî pa = ¼ and the two elements in blue give â-na.

Speaking of annas, the purple Persian in the center of the rectangulars is just that. Its dotted tooth represents the single ‘n’ (we do not know why the ‘n’ in the usual English transcription came to be doubled.) There are two common spellings seen: the one shown here has the so-called choti he at the end (an ‘h’ not pronounced here.) The word is also often rendered ânâ with a long-final, which is more Indian, less Arabic, in flavor. The initial vertical stroke alef is really a vehicle for introducing a number of different initial vowels. Here we have long-â by virtue of the twiddle, called madd(a) on top, which is secretly just another alef written small and sideways in an aesthetic tradition. One claim is that two parallel alefs in a row would be “most hideous.” The madda often goes missing in practice.

And speaking of vav, here it is again in yellow. In isolation, as here, it means “and.” It is pronounced in Persian either as va (parallel to Arabic wa with a counterpart in Hebrew) or as o, an independent development in Persian for combining closely paired nominals, as in the case at hand: Jammûn o Kashmir. As always, “red for Jammu, blue for Kashmir,” which follows the tradition that was more or less adopted by the stamp printers as well.

In shape, the vav is really a loop, but it often comes filled-in or unclosed, and so can be mistaken for several other letters in cursive writing. The letter is mostly seen in vowel use, especially serving as long-û, as we may see already in the Jammûn just above; we’ll see it again below in mahsûl ~ postage (tax) on the New Rectangulars, and yet again serving as a short-o when we look at the 2-anna stamps. And then there is that troublesome qalamrao, qalamrav, ..., for which the ending is variously rendered under dialect influence. Similarly, the pâv ~ ¼ is sometimes understood as pâ'o, i.e., with the glottal stop or hamza interposed. Seems it links to Sanskrit pâda.

½ anna. The yellow is Persian nîm ~ half. The dotted element n is immediately attached to the m-blob, written here in its final form having the long downward tail. The internal vowel, though long, is not written here. The pair of elements in red is Dogrî a-da ~ half. Properly it is rendered adha involving a different script element.

1 anna. The yellow is Persian yek, where the two dots under the initial tooth signify y-. The short internal vowel is not rendered, and the long horizontal is the flourish on final -k. The Dogrî correspondent in red is ek, similar to the Urdû pronunciation.

By the way, in the lower section of this strange 1a seal is an example of the ek in the Urdû script. Notice the introductory vertical stroke, the alef hamza, that destroys the palatization of the Persian, if it may be so put.

2 anna. Persian do ~ two. The first yellow comma-like element on the right is Persian d. The other is the vav mention above, serving here for short-o. The red Dogrî element is do (~dô), where the Dogrî diacritic indicating the -o is also missing.

4 annas. The yellow is Persian chahâr ~ four, and the red is “Dogri” chahar, but it is really just transcribing the Persian, for the Dogrî word is châr.

8 annas. The yellow is Persian hasht ~ eight. The initial h is the blobbed hook and the three yellow dots are marks for the sh. Without them, the letter would formally be s. Often in handwriting the dots are missing even when sh is intended. One might expect a pair of upper dots on the final flourish of the t. While more clearly printed stamps might reveal them, we suspect they are missing. The red is Dogrî ât' ~ eight.

Inscriptions (New Rectangulars)

In the New Rectangulars, the denominations can be found as spelt-out words in Persian at the bottom of the central region, while the denominations in Dogrî are situated between the two stars at the top in the inscription band. In the case of the 4a and 8a, which was engraved by a different seal-cutter, this Dogrî at the top has the bottom of the lettering facing outward, and the two stars have disappeared.

The rest of the Dogri in the band, which is repeated on all the stamps, merely repeats the Persian from the center. With the bottom of the script facing outward, read counterclockwise: mahsûl dâk kalamrao jama o kâshmîr ~ postal tax of the Dominion of Jammu & Kashmir.

Here for comparison with the preceding is the Dogrî ‘kalamro’ and ‘kalamrav’ from the postcard and from the 8a pane. The last image shows a constant error in position #7 from that pane: spy the missing -la-. There is an unclever watercolor forgery of this in which the error is corrected. General notice of this error is due to Frits Staal.

The Urdû bâlâ that occupies the rivet positions at the top of the 1-anna and 2-anna New Rectangular sheets means just that: ‘top’ of sheet.

1/8 anna. nîm pâv âna ~ half of one-quarter anna, in the Persian. The Dogrî script at the top is merely transcribing the Persian, not rendering the Dogrî itself, which would be âdhâ pâ ânâ, though not all these long vowels would generally be made explicit.

¼ anna. pâv âna in the Persian. In the Dogrî at the top, the diacritic after the pa-element indicates long-a. It is meant to be a little v-shaped mark. It is seen more clearly in the forms for ânâ used here and in the anâ of the next image. One really expects ânâ, but such details seemed of little import.

½ anna. nîm âna in the Persian. Dogrî at top is adha anâ, and this time is not a mere transcription of the Persian. This example is good for inspecting the rest of the Dogrî in the band, which repeats the content of the Persian.

1 anna. yek âna in the Persian, where the pair of underdots is shifted abnormally to the left. (Plate postion #8 omits the dots altogether.) Dogrî at top is formally yak ânâ, again merely transcribing the Persian; otherwise we would see the form ek as in the 1a Old Rectangular above.

2 annas. do âna in the Persian. Dogrî at top is do ânâ. The diacritic for Dogrî o is an upside-down v-shaped mark that is often placed above, not after, the syllable element in question.

4 annas. chahâr âna in the Persian. Dogrî at the upper-right is char ana, i.e., not merely a transcription of the Persian. The form seen here is like nagari-r, such hybrid writing being common. The distinctive Dogrî form is like a partial derivative sign, as may be seen on the 4a Old Rectangular above.

§2 Paper

A modern and important discussion of the papers is carried out by Tim Eames in India Post 28, 49-52 (1994). An older discussion by Séfi & Mortimer is accessible here on-site†, which includes a summary of Evans’ 1903 treatment.

Native Paper

Watercolor stamps appear on a sturdy, thick striated paper called native, local, or Indian paper. Our pet name for it is papyrus. It is not the same as the India paper, also called China paper, mentioned in older philatelic literature. Thicker samples of the paper can be somewhat springy and it is relatively resistant to tearing. Though the striations may on occasion be a little difficult to make out, it is never difficult confuse this paper with the competition, namely European laid papers and a variety of woves.

Native paper thickness: The thickest section we have measured is ~ 0.26 mm and it will certainly come thicker. By contrast, the thinnest of the thin woves approaches a tenth of this value. Nothing about chronology is safely asserted from the evidence of native paper thickness since a single sheet may show wide variation across it, as sample sheets in Masson’s book attest. Paper thicknesses are important, however, among the New Rectangulars.

By the way, we definitely recommend Precision Graphic Instruments, Inc.† as a supplier for the North American market of analog and digital micrometers for measuring paper thicknesses. An uncertainty range of 0.01 mm (a 10 micron range) is satisfactory.

The Native Paper-Making Process

For some interesting details about the paper-making process, do check out the section in Staal’s book. Masson notes that the paper manufacture in Kashmir at the time of his writing was carried on in the tahsîl Naoshahr (on the road leading from Srinagar to Ganderbal, he says) and also in the State Central Jail at Srinagar.

European Laid Papers

European laid paper is first seen postally in 1877. It tends to be thinner, smoother, and less springy than the native paper. The lines are more regular and more evenly spaced lines than those seen in the bumpier striations of the native papers.

The thickness of European laid paper actually matters. On the left is an item from 1878 that was printed on “medium” European laid paper, which comes in the range 0.10-0.12 mm. Such a thickness stands in important contrast to the much thinner paper that appeared a decade later in the ~ 0.05 mm range (item right) usually known as ‘creamy laid’ paper. That designation is used for reasons here unknown. Perhaps this paper has just greatly yellowed over the many decades since the nomenclature was fixed. (Or maybe cream from the Kashmir ibex is yellow? And I don’t want any.) These latter stamps will all be from plates showing late plate-state characteristics, witness the screw impression in the border, whereas the medium laids will be in the earliest states of the plate, i.e., no screwhead marks. More detail about the New Rectangular plate states is given at the bottom of the screen.

Above: A range of laid papers are also found among period forgeries and later. A notable type is this very soft, white laid on a mirror reversed forgery. Collection Lunn.

Wove Papers

Wove papers, also introduced experimentally in 1877, come in a broad range of toning, coarseness, and thickness. Thin papers appear in the 0.05 mm range, mediums in the 0.10 mm range, and thick papers (most uncommon) come greater than 0.12 mm. The only known wove among the issued circulars is a rare thick, coarse, brownish wove, known in the literature as “sugar-wove”. It was not used for any New Rectangular. It is known from a few singles and a half-dozen covers in April 1878.

Some of the thin woves are yellow-toned, as here, others are grey-toned. There is a marked range of smoothness, fine to rough, in both tone types. Except for the tinted varieties of paper (shown downscreen a bit) such distinctions of toning and character are not distinguished in the mainstream catalogs, though the paper differences alone obviously make for some starkly diverse material, some of which is much scarcer than other.

The late bright white variety of smooth wove paper known from 1889. It is also found in some of the late forgeries, as seen in the missing-die forgery of the 1a circular, which is presumed to date from the same period.

Above: An example of the grey-toned thin wove paper. This item was printed at Srinagar in the early 1880s, and some of that stock was re-issued in the ‘90s (i.e., not reprinted.)

Tinted Woves

The tinted woves are confined to the 2a values of the New Colors regime that began in 1883. The colors (being reds in this case) are not exactly “new”, but the papers made the stamps something of a novelty at the time.

On the left is yellow-tinted thin wove. This specimen is in a very thin (~ 0.04 mm), fine, and almost transluscent paper, like certain smooth tissue papers. The pigments used sometimes seem rather thick and oily, and invariably of a orange-red hue. There is virtually a continuum of these tints from the purer yellows, through yellow-greens (as shown on the right) to the outright greens shown next. As to dating, the yellow-greens may have been the earliest type, with purer yellows appearing late.

The green wove papers are decidedly scarcer than the yellower papers and come in a range of thicknesses, from remarkably flimsy indeed (~ 0.03 mm) to thin (~ 0.05 mm.) The tinting makes even more difficult than usual any judging of the shades of the red pigments. As to datings, there is a discussion in Séfi & Mortimer in this on-site link†.

This specimen shows distinctive printings of the 2a on a thicker (~ 0.08 mm), coarse, opaque version of the yellow-tinted paper. These are known on cover in the 1892-94 period.

Pelure Papers

According to the Stamp Collector’s Encyclopaedia, the papers known as pelure (from Latin pellis ~ skin) are strong, thin, translucent papers with a barely perceptible wove or laid pattern. Any printing clearly shows through from the back. Such papers are known in some of the experimental printings of 1877. The tinted woves mentioned above are sometimes called ‘semi-pelure’. They are not so tough as full-fledged pelures, and their transparency owes as much to their thinness as to an inherent translucence.

Bâtonné Papers

Bâtonné paper comes in both laid and wove forms (and Poonch boasts a ‘ribbed’ type.) It is characterized by a set of prominent and widely-spaced parallel lines superposed on the basic wove or laid pattern; in the case of the latter these lie either parallel to or perpendicular to the much more narrowly-spaced laiding lines. A type of smooth writing paper known as “cream laid” is often bâtonné. Quadrille paper bears a two-way bâtonné pattern, but such is not known in the Kashmir material.

§3 Papermakers’ Embossments

Animal and date embossments are occasionally found in the paper of the New Rectangular stamps (excluding the very thin tinted wove papers, which may not bear them successfully.) These were done by the manufacturers of the paper and as such are not really part of the stamp production process itself. They are reported over the dozen-year period 1879-91 on thin toned woves, the later thin laid papers, and the late bright white. In our sampling, about one in 300 New Rectangulars (not counting the tinted papers) show these markings, which would imply that not all foolscap-sized pages (with four stamp sheets to a page) were so treated. The embossment ovals invariably occur on the corner stamps of a sheet, both erect and inverted with respect to the stamp impression. As it happens, earlier quarter-annas predominate in our sample, but all denominations do occur. Shown left is the “small elephant” control mark in an oval ~ 14-by-19 mm.

Above: Englargement of the “large elephant” mark in an oval. The actual size is 16-by-22 mm. There is a rider, house (‘howdah’), perhaps a small village, on the poor beast. Below is another glimpse of an elephant in the large-oval format:

A rare unicorn has also been sighted (ref Staal, p 115, reporting on the discovery by Dan Walker.) This specimen was offered in the Haverbeck auction (Lot 1478) as existing on the ½a red on thin wove paper, on the top left corner of the plate, and thus exhibiting explicitly the state I condition of the plate. There is also rumor of an elusive prancing horse (assuming it is not a unicorn.) It is reported unframed on a ½a red on thin wove paper.

The earliest control date is 1877, i.e., on paper manufactured at least a year before the advent of the New Rectangulars themselves! This embossment was upside-down in the lower-right corner of a sheet of ¼a reds on thin wove paper, which is an 1879 Jammu printing. The year 1877 was in fact the date of the stamp-printing establishment at Jammu of the new Ranbir Prakash press. It is interesting that thin wove paper was available at such an early time, a time when paper experiments had just started on Old Period stamps with European laids. Thin woves are indeed known (“anomalously”) during the early New Rectangular regime; more surprising is that there were not more of them, since the paper stock was long available and the thin woves were ultimately to triumph.

The embossment ovals containing year dates attested by Séfi & Mortimer are 1877, 1878, 1879, 1881, 1882, 1883, 1884, 1888, and 1891. Missing amongst these are thus 1880, 1885, 1886, 1887, 1889, and 1890. We report an 1885 herewith. An 8a official in the thin laid paper been seen with an 1884 date embossment, which seems early indeed for these ‘Prataps’. Again, older paper stock was presumably being used in a later printing of the stamps. Still this encourages one to be on the alert for Prataps in pre-1887 postal usage.

First reporting of 1885? This control mark appears on a ¼a black official sheet on coarse, yellow-toned thin wove paper.

§4 Perforation

Certain printings in the 1878 New Rectangulars period are known perforated. Most are very rare (and/or doubtful) with only the ½a reds on laid paper at all common. A harrow-like implement consisting of sharpened brass-pins in a configuration suitable for perforating the 15-subject plates has actually been seen by philatelists. It provided rough perforation in the 10-12 range.

That a second implement usable with the 20-subject plates also existed is shown by the detail above, which was taken from the unique watercolor (!) proof sheet on laid paper of the 2a design (with thanks to Wolfgang Hellrigl for the image.) The same exists in the 1a, also in the Hellrigl collection. These do not seem to be of significantly finer perf pitch than that found for half-integral issues. The traditional division into wide pitch in the 10-12 range for half-integral stamps and a fine pitch in the 13-16 range for the integral-valued stamps has long been questioned.

Perforated New Rectangulars

Late experiments in may have had official sanction. A 1/8a yellow-brown, ½a orange-red, and a shades of 1a greens are reported in clean perf 12 on thin wove paper. These are also reported as having been gummed. Unfortunately, though not surprisingly, doubtful specimens are afoot. A block of nine 1a “bright green or bronze green” on a large piece of a registered cover to the Raja of Jammu and dated 18 December 1893 was so described in the Haverbeck auction catalogue. There is also Dawson’s account in The Philatelic Journal of India 43 122 (1939). Staal speaks of the same item as dated instead 1891 and in the shade dull green.

§5 Watermarking

We refer the reader to the Séfi & Mortimer watermark† link on-site, together with our regrets for the absence of their drawings. Though much of the European laid paper of the old period and of the early New Rectangulars is known variously watermarked, watermarks have played little role in J&K philately, whether for dating or for the identification of separate printings. Watermarks were done in large-format designs on large sheets, so one may expect an occasional watermark line or two tracing across a stamp. Some of the watermarked paper known in certain Afghanistan issues is the same as that known in certain 1878-79 Kashmir issues.

§6 New Rectangular Plate States

Four of the New Rectangular plates (¼a, 1a, 2a, 4a+8a) are traditionally divided into two starkly separated conditions known as “State I” and “State II,” the latter characterized by screw-head impressions in the margins of printed sheets. The received story is that these appeared when the plates were rebedded after their removal from Jammu to Srinagar in 1881. The little-used & late 1/8-anna plate (1883 Srinagar) never suffered such treatment, and its slightly evolving condition over time has somehow elicited little in the way of public fascination. Finally, there is the rather interesting case of the much-used ½a-plate, which had a sequence of more complicated repairs with rivets, not screws. The final state of all the plates was cruel defacement.

Since the plates were in heavy (and evidently not-so-careful) use for a dozen years or more after the screws and new rivets were employed, finer discriminations of plate state can in principle be done, perhaps even usefully, for helping to constrain dating sequences of the many shades.

• Higher-resolution scans of all the New Rectangular plates, which can be used for plating work, can be found on-site in this Appendix† to the second Masson volume.

The Five ½a-Plate States

The account of the ½a plate is more involved on account of a broken section along the top of the plate and rivet impressions that appear and disappear over the course of time. What follows are five scans that follow the history of this plate:

• State I. The earliest printing at Jammu is characterized by a white crack line that crosses the top of the plate. Just above the crack line is an example of a rivet hole containing a well-executed rivet. Do keep an eye on the rivet impression near the lower-right corner, for it figures heavily in the rest of the discussion.

• State II. The upper segment of the plate above the erstwhile crack goes missing and so too a number of rivets, leaving holes. New rivets, not so well executed, have also been reintroduced into the four corners, evidently to rebed the plate to a new base. (These will again prove insufficient.) Whether this repair work was actually done at Jammu on account of the crack problem or later at Srinagar is not obvious and the matter invites a closer study of the dated postal material. Séfi surmises that the missing segment at the top was removed during the disbedding of the plate when it was being prepared for the removal to Srinagar, and this new state represents the condition of the plate shortly after its rebedding there with the four new rivets. It is our duty to be on the lookout for this state of the plate on earlier-dated material.

• State III (rare). This state is characterized in particular by the presence of a rivet impression in the lower right margin that was missing in State II and will go missing again in State IV. The scan shows that this rivet somehow caused the paper to tear when the sheet was removed from the plate surface. Perhaps this or some other nuisance about this rivet was deemed troublesome enough to warrant its forthwith removal. A number of other rivets have also been added, and the quality of the work has continued to deteriorate. The new rivet heads evidently projected slightly above the plate surface; for in that case the large ink-free annular regions around each head can be accounted for as a result of the heads depressing the paper surface. Scan from Geoffrey Flack. See also the discussion by George Harell in India Post 39, 129-32 (2005).

The presence of the white disc in the vertical black band on the left side of this image is an un-inked version of the rivet feature that characterizes State III. Keeping the rivet-head un-inked perhaps helped to keep the paper from tearing there. This scan, taken from Plate 31 of Séfi & Mortimer, is from a simultaneous printing of the ½-anna plate together with a State II 1-anna plate (as indicated by the screw-head impression in the floral margin.) The existence of this image was fortuitous in that it revealed the existence of this rare plate-state III of the ½a plate to those of us who had never seen an actual example.

It is interesting that Séfi & Mortimer’s other presentations of tandem printings involving the half-anna plate show it to lack the rivet in question, already or still. Those authors and the other older literature do not mention this rare state of the ½a plate, and so distinguish only three basic types, their State III being perforce our State IV, and our State III being special.

• State IV. Notice that the rivet in the lower-right corner is gone (again), leaving a hole in that position, as in State II. This scan shows the condition of the plate in the late period, an impression on the thin laid paper (>1887.) This plate state is by far the most commonly seen, and was printed in the black on smooth, thin, white paper in huge abundance in the later period, possibly for dealers & collectors, and thus act much like reprints. Caution: Recently there have appeared also rather good photocopies of such.

• State V. A black & white scan of one of the Staal-Sharma purple-ink impressions made in August 1981 from the defaced ½a plate. This impression shows that the rivets could indeed be removed successfully, returning the margins of the plate essentially to State II conditions, apart from the four corner rivets, which are now also missing.

While we are at it, there are a couple of other curiosities of the ½a plate. On freshly-cleaned plates one may see a style of so-called “Maltese” cross in the left margin, just beside subject 4. Though it does show up on some rather early State I sheets, we do not know if it is a feature from the very beginning of the production. Brighton forgeries of this plate, which come in ½a brown-lake & scarlet on medium wove paper, are said to show the cross very clearly—too clearly and too large in fact, all in evidence of an ill-conceived modification of the plate by the forgers. Should we ever come across an example, we promise to put its image in this section.

Some authors refer to the disfiguring scratches as seen above in the left scan. Such are in variable evidence depending on the heaviness of the impression and on how well the plates had been cleaned. Shown here are early and late plate states showing the here-today gone-tomorrow phantom scratches. In heavy inkings even the cross may not appear.

Numbers are scraped in the margin of this corner specimen. You will discover that an attempt to inscribe numbers with this appearance on sheets today will fail. One concludes that the operation was done at the time of printing when the ink was still wet. But no actual etching of the plate ever took place the way it had been for the cross.

To top of page