Re-Entries

Re-entries and repairs.

Recent publications have illustrated the fact that understanding of the different types and effects of re-entries are not universally held by all philatelists. In fact it probably varies person by person. In addition, I would like to expand the accepted types of re-entry since I believe that there is a need to do so.

What is a re-entry? It is when an impression on the printing plate is subjected to a subsequent attempt(s) at entering an impression from a transfer roller. There are various reasons for attempting this and differing results, as follows.

Shifts

Starting with a phenomenon that is not a re-entry is perhaps an unusual way of getting the ball rolling, but the fact is that these can look very much like re-entries. The normal type of shift is what is called a basal shift and shows itself by a doubling of the bottom of the stamp. Occasionally one can find doubling at the top of the stamp, that is - a top shift. The cause of this feature is the transfer roller spindle slipping within its holding brackets during the transfer process. This problem was solved (Bacons Vol 1 top of page 58) [When? Gibbons lists a couple of examples on plate 84, which was registered October 1848] by the simple expedient of introducing a grub screw within the bracket that could be tightened against the transfer roller spindle. [egs plate 17]

Stone states that it is caused by a wave of metal riding before the roller, taking with it the impression of the base (or top) of the stamp. This seems to make sense, until you realise that it was solved by the above mentioned screw.

Additional note. This cause is also stated within Fundamentals of Philately (1971). I have been thinking on this and it does make sense up to a point. Elsewhere (see below), I describe how metal reacts when put under pressure. Under slight pressure metal exhibits elastic deformation, which as the word implies, is a temporary state. More pressure results in a plastic reaction, which has some permanent effects, and would be the goal of the engraver. Applying yet more pressure would result in necking, where the metal under pressure not only suffers damage through metal fatigue, but also hardens as its structure as it is compressed. I think that this last stage could result in the metal being pushed (shifted) down the surface of the plate, but only when the process was rushed. A steady increase of engraving depth over a period of time would not result in necking, but only of plastic deformation. The engravers at Perkins Bacon, who had a good deal of experience working with steel, knew that to rush the job was to invite trouble. Remembering again that the problem stopped after the fitting of the grub screw shows the truth of this. A more inexperienced company could have been forgiven for rushing the job and in the process creating shifted transfers as described by Stone and others.

Fresh entries / First states and Re-entries / second states.

Here it is important to make our definitions clear. “Fresh entries” are those impressions that have, for one reason or another, been repaired by the re-entry of the impression on the transfer roller before the plate has been put to press. It follows that these stamps will be present on the imprimatur sheet. Stamps that are produced by by the plate in this state are said to be in “First State”.

If a reapplication of the transfer roller is made to an impression on the plate after the plate has been at press, stamps from that impression are called “Re-entries” and is said to be in the “Second State”. Subsequent reapplications of the transfer roller will produce “Third States” and so on.

It is important to note that it is not just the transfer roller that can produce later states of an impression. As an example, if an impression suffers from wear after a period of use, part of the design on the plate can be restored by using hand tools. The result is not a re-entry, but is none the less a second state.

All “First State” roller re-entries should be referred to as “Fresh Entries”. All “Later State” roller entries should be referred to as “Re-entries”. Stanley Gibbons – please note!

Types of roller entry

First entries

This is a term that I have made up to define the impression made by the transfer roller on a blank, unhardened steel plate. The vast majority of stamps are produced from such an impression.

Fresh entries

When the engraver positions the roller out of line with the surrounding first entries, he is left with a problem. If the resultant impression is so far out of kilter that it would encroach into the design of neighbouring impressions, he has little option but to try to remove as much of the faulty impression as possible. Once he had done what he can, he then proceeds to re-enter the impression in the correct position. In many cases it was difficult to remove all traces of the previous impression. Thus we see a doubling of the image. These type of re-entries produce some of the most outstanding examples of any type of re-entry such as the 'Union Jack' re-entry of plate 75.

[Adjustment entries

In some instances, it is probable that the engraver spotted the inaccurate positioning of the roller after the initial contact of the roller with the plate. If the depth of the impression made was not very great, it would have been possible for the position of the roller to be adjusted and process restarted. This would have left a shadow entry that to all intents and purposes would have the same results as a fresh entry described above. The only difference between the two types of re-entry would have been the attempt to burnish out the old impression or not. As far as looks, it would be impossible to tell apart fresh entries and adjustment entries, so the existence or otherwise of adjustment entries is a moot point.]

Hand repairs

One the plate has had it's full compliment of 240 impressions, the next job would be to punch the check letters into the plate. We are now left with a plate that would be impossible to print from. Around each impression would be small mounds of displaced metal (burrs)and around each letter, the same. In order to print from a plate, the plates surface needs to be absolutely smooth and level, it follows that the removal of these burrs is the essential next step. Generally speaking, this operation was done with skill so that few problems arising. Occasionally, we find what I would call a burr rub, that is, an area within the impression where printed detail has been reduced due to the over exuberant use of the burin. As an example see plate 76 FH-FL, where three of the impressions show broadening of the word Postage.

It is not unusual to find areas on the plate where a small amount of ink tends to remain after wiping, especially in the vertical gutters between impression. These are often referred to as burr rubs, and I suppose they are. However, in an effort to try to be specific, I would be more inclined to call them simply 'blurrs'. The matter gets worse, when we come to re-entries that result in what I would call Transfer Roller Lines, and some others call 'scrapes', 'burr rubs' or simply 'rubs'. Personally, although I understand that the cause of the appearance has something in common with each orther, I would prefer to limit the phrase transfer roller line to second states, even if the bloody thing does look like a blurr.

Post hardening fresh entries.

The hardening process is one that is fraught with the potential to damage the newly engraved plate. The temperature necessary for hardening is not too far off the point where the steel will start to soften with any slight error of judgement having possibly damaging results. We have the engraving book covering 1863 onwards, and occasional mention is made of impressions being damaged during the hardening process. For example, the entry for 24^th February 1864 mentions “Postage Plate No.82 [Die 2] 45 m[inutes]: Fire fierce and unequal. 4 heads injured.”. It is inconceivable that they would have put the plate through a softening process, re-entered the damaged impressions and then re-hardened the plate. The damaged impressions would have been re-entered without softening before the imprimatur sheet was printed. It may be that evidence of such a re-entry would not be apparent. However in some instances, I believe that that it may be possible to show that some fresh entries can be described as post hardening fresh entries.

Once all of the above had been completed and the plate cleaned, the plate would have been used to print an imprimatur sheet. This effectively draws a line in that plate's history. Which is nice.

Re-entries

Re-entry periods of activity

Early die 1.

Another flurry around 53-54 including the unhardened experiment. More to do with the shallow die 1 than anything else.

Plate 33 in early 1857 and 38, 45, 55 and 56 in late 1857

then again in 1861, with plates 48,51 and 52.

Another batch around jan/feb 68, including plates 73, 80, 81, 85, 90, 96 and 100. Some of these are in the engraving book, though recording is inconsistent.

The availability of good quality steel plates may also have been a concern. The right quality was difficult to get hold of. As more plates stayed at press for longer, there would have been a need that may have been difficult to fill.

Method

The plate was removed from the press. If it was intended that printing was to continue with the plate, the packing that had been arranged on the printing press would have been carefully left as it was. This would not have been important if the plate was coming off the press for repair and after put into reserve.

The plate was placed upon the smaller transfer press and the plate and roller die were aligned with each other. It will be remembered that in most cases before the introduction of the Holtzapffel ratchet equipment and narrower roller dies, most impressions on a printing plate were slightly askew. Probably the way to get around this would have been to fix the plate to the bed of the transfer press, and then position the loose transfer roller by hand until it engaged with an impression on the plate. With the roller still floating loose and under no pressure, it would need to have been rolled along the impression on the plate. If it was smooth, the alignment was correct, and the nuts or screws that fixed the alignment of the transfer roller were tightened. Hopefully, the impressions on the plate would have been aligned similarly, especially if the whole plate was laid down in the same working day. Though this is not a given.

The next thing that would have been for the engraver to fix the depth of the roller, that it the maximum amount of movement allowed vertically. This would have been easily achieved by laying a blank part of the roller on the plate surface, and tightening the stop screws at that point. It was essential that the surface of the printing plate was perfectly smooth as wiping excess ink or indeed printing would have become impossible. Going too deep into the plate surface would have produced a trench, which would have effectively written off the plate.

Once aligned, the engraver could start the process of rolling in the impression, that is rocking the roller over the impression repeatedly. As he did so, he would lower the roller very slightly, under pressure. This would have the effect of deepening and hopefully repairing the worn or damaged impression.

On hard or softened plates?

It seems to be generally accepted that before plates were re-entered, they were first softened in a furnace, re-entered and then re-hardened in a furnace. I think this is based on a general belief that to try to re-enter a hard plate would effectively destroy the transfer roller. There is probably an element of truth in this as the number of impressions that are repaired on any single plate do tend to be limited. Compare with plates that were left unhardened and had all 240 impressions totally re-entered.

What is overlooked is the fact that trying to soften a plate was to risk destroying it completely. The higher the carbon content of steel, the lower it's melting point. As a general principal, most case hardened printing plates would have a higher concentration of high carbon steel precisely at the most delicate points: the raised, fine details on the plate surface. To try to soften a plate, was to risk turning the Queen's image into marshmallow blobs.

Bacon [Vol 1, pg 60]states:

“...the impressions of the stamps, where the hardening had worn off, would soon show greater signs of wear than those on other parts of the plate. When this happened, the worn impressions would be " re-entered " without softening the plate, and after this had been done the plate would often be left as it was: at other times it would be hardened a second time.”

Softening or hardening would have taken at least a day to achieve, possibly two. So to completely repair a plate would have meant that it was taken from press for a minimum of three days and maybe up to a week.

What forces are involved here?

Have a look at Wikipedia or the suchlike under “Deformation” for more information about the stresses that would have been inflicted by the re-application of a transfer roller. Here is an oversimplified summary.

Elastic deformation. Where enough stress is applied to in this case a metal to bend it, but not beyond the point where it cannot bounce back. Slightly bending a wire coat hanger can illustrate this very well.

Plastic deformation. More stress is applied to the metal which takes it beyond elastic deformation, and a permanent bend is the result. Give that coat hanger more of a bend, and it will bounce back part of the way. But not all of the way. This an interesting stage as squidging the crystalline structure will cause a hardening of the area under stress – have a look at “cold working”. This is followed by what is called necking, which is the start of the breakdown of the structure of the metal which will lead eventually to fracture.

How did the metal of the PB plates react? Well that depends on the actual structure of the steel being examined. If we take Mr Grazow's view that there was a large proportion of martensite in the case-hardening, then attempting to re-enter the plate without first softening would have been disastrous. Perhaps a bit like trying to mould cold glass. End of plate IMO. Other structures? Ferrite – no problem at all as it is very ductile. Cementite – this would have been more of a problem as it is very hard and brittle but possible if it were embedded in more ductile structures. Pearlite – would probably react reasonably well as it has a laminar structure of cementite and ferrite. Bainite – very hard, but with a granular structure within ductile ferrite. Probably a good mix to try a re-entry into.

The engraver at PB would have been largely in the dark as to how a plate was going to react to being re-entered until he started the process. Slowly does it would have been the way to go, and listen to the noises coming form the work. Silence – that's good. “Graunch, scrunch” - Oh dear.

The main stress axis would have been from top to bottom of an impression. A roller in co-incident contact with the plate would have nestled snugly in an impression from left to right. Lowering the roller would not have altered this, but would have introduced stresses in the metal above and below this plain. The stresses described above would have been seen vertically, above and below the point of contact of the roller.

Signs of re-entry

Transfer Roller Lines (TRLs)

There is a separate section on this web site that covers TRLs, though I have just noticed that it is far from finished. The key point to try to decide is to date the marginal blur. If it is present on the Imprimatur, then it is either the result of over zealous removal of a metal burr, or possibly the result of a fresh-entry. If there is no blur evident on the imprimatur but another copy of the stamp emerges with what I would call a TRL, then as far as I am concerned it will be the result of a re-entry of the transfer roller. I cannot think of any other way for it to get there.

Strengthened sides / background / facial details

This is one of the most common effects of re-entry, which is perhaps not surprising since it is after all the aim of the re-entry process. Comparison of as many examples of an impression as possible is the cornerstone of a study here. I think the trick is to build a timeline for the impression. One has to be careful not to confuse the effects of wear and re-entry. For example, an impression may not be a repaired impression just because it is a stronger print than another. It may just be an earlier print. Other clues should exist to point the correct conclusion. Some knowledge of different ink colours and of course dated examples will help get the timeline right.

Compressed “Postage” and “One Penny”

Robert Folkard was fond of this phenomenon and described many stamps from 173 as having it. I have a problem with it as I find it far from simple to see a difference in many cases. But it certainly can exist as a look at the states of plate 12 TL will confirm. I think it is more a product of a slightly non co-incident re-entry than anything else.

Weakened check letters and hardened letter cages

Check letters, if struck soundly, would be one of the deeper parts of a stamp's impression. Thinning check letters from re-entry can be found on many plates, but it is far more common on some than others. For example, on plate 155 and 176, thinning check letters are a good indication that the stamp is from a repaired plate. On 173, the letters are perhaps more often found re-cut, which would indicate that they were very weak. Weak or re-cut letters are also found on plates 1 and 5. The key point about these plates is that repairs took place on unhardened plates in my opinion.

If a re-entry is attempted on a hard plate, the case-hardening would still be entirely intact in and around the depression that made up the check letter. It would have formed a hard cage that would have resisted the transfer roller. The line of least resistance would have been the softer ferrite mix below the surface of the plate, and the cage would have been pushed into this metal, possibly without damage. On unhardened plates, there would be no case-hardening to preserve the shape of the check letter, which is why we see them thinning.

Wooly appearance

This is just as it says. Speaking to others this seems to be personal to me, so is almost certainly of nil value to the average reader. Sorry about that. When looking at a stamp from a freshly repaired plate I can sometimes discern a feint wooly appearance in the printing. I think it has something to do with the contrast between the background lattice work and the rest of the stamp. As the plate wears this appearance fades, so it is not a constant feature that can be relied upon. Plate 155 seems to show it most commonly for me. Of course, I could be off my rocker.

AP

July 2014 - Dec 2017.