Every folio of the Voynich MS is numbered, in the upper right corner of each folio. In the case of the large foldout folios, the number was added when the folio was completely folded in. A reasonable first assumption is that the folio numbers were added to the bound book, but we will come back to this question elsewhere, in another context (see below).
Quire numbers (mostly Latin ordinal numbers) are written mostly in the lower right corner of the verso side of the last folio of each quire. D'Imperio already mentioned the difference in style between these two sets of numbers, and this point is further analysed in Pelling (2006) (1).
John Manly suggested that the style of the quire numbers belongs to the 15th Century (Manly (1931) (2)). The folio numbers appear in a style that is less constrained in time. De Ricci (1937) (3) gives the 16th Century and Manly (1931) (see note 2) the 17th (based on the handwriting style). A 15th century date may also not be excluded.
What seems certain is that quire and folio numbers must have been added by different people, even if we cannot say how long after each other. The numbers are also fully consistent with each other (though with one curious exception that is mentioned in another section below). Both series increase monotonously throughout the manuscript.
To complicate matters, Nick Pelling proposes that several different hands in the quire marks can be identified (see note 1). What is certain is that the quire numbers on quires 19 and 20 are different in that they are not written as ordinal numbers. Quire number 20 is also not written in the usual location (verso of last folio), but on the front of the first folio of the quire.
The two missing quires (numbers 16 and 18) coincide with gaps in the foliation (2 missing folios each). This tells us that the persons who added these sets of numbers must have been fully aware that there were other folios that are now missing. The most logical assumption is that they were not missing at the time. How else could the foliator have known which ones were missing?
Another interesting observation follows from quire 8. It is the first incomplete quire in the MS, and it consists of only two bifolios instead of the expected four. However, the folio numbers jump from 58 to 65, showing that there should have been three bifolios in the now missing part, meaning that this quire would originally have had five bifolios. This cannot be the result of a guess. The foliator must have known that there were three folios missing. These folios must have been there, and numbered, some time in the past.
There are several reasons to believe that the current page order in the Voynich MS is different from the original order, or at least the originally planned order. While some, if not most, of these arguments are quite sensible, they are difficult to prove as long as we are not able to read the text. Following are the most obvious cases where the present page order appears incorrect in one way or another.
After a long herbal section, there are astronomical, cosmological and biological sections, only to have another number of herbal pages which are not fundamentally different from the earlier ones. They look out of place, as if they should have belonged in the earlier part. Also, in the early herbal section, there are a number of bifolios written in a different hand and with different text statistics (4). The text statistics have been referred to as 'language A' and 'language B' by Prescott Currier, and one can therefore also speak of the two different types of herbal pages as Herbal-A and Herbal-B pages. The bifolios in these two styles seem arbitrarily mixed, as if it was unintentional. What has not yet been clearly established is, whether the drawing style of Herbal-A vs. Herbal-B pages shows any consistent difference. The tendency is that this is not the case. If true, that would suggest that the drawing of the plants and the writing of the text were largely independent activities. It is also to be noted that both the foliation and the quire numbers appears on herbal-A and herbal-B pages alike, in the normal order.
In the biological section, the bifolio consisting of folios 78 (verso) and 81 (recto) together form an integrated design with water flowing from one folio to the other. This design would have been visible if the bifolio were in the centre of the quire, but it is not (5). Furthermore, Glen Claston has observed that the biological section could be considered to consist of two different topics, and should have formed two separate quires instead of one. This is discussed at the Cipher Mysteries web site. (to be added). This argument is a bit more speculative.
Quire 9 consists of a single very wide foldout of folios 67 and 68. Folio 67 is two panels wide, while folio 68 is three panels wide. It has long been noted that the quire number (9) appears to be in the wrong place. At one point, John Grove noticed that there appear to be stitching gaps of a previous binding in one of the foldout creases. By assumung that this should have been the original binding crease, all of a sudden the quire number falls in its right place. This is very clearly described in Pelling (2006) (6).
While the quire number is now in the wrong place, the folio numbers 67 and 68 are in the right place for the present binding, yet would be in the wrong place if the binding had been made using the unused fold. What really happened with this bifolio requires some further thought, which is being addressed elsewhere.
The pharmaceutical section, which consists of 3 bifolios that should belong together, are spread over two quires (quire 15 and quire 19) with some herbal folios in between. In addition, based on the style of the container drawings, the two bifolios of quire 19 logically belong before the bifolio in quire 15.
From all of the above, we may conclude with high confidence that the present order of the folios in the MS is not the originally planned one. It has not yet been possible to reconstruct the most likely original order, and it is also not entirely clear how many pages are presently missing (beside the 14 missing numbers in the foliation).
In the following, all observations that can potentially contribute to our understanding of the order in which the MS was constructued will be summarised one by one. At the end, a tentative conclusion will be drawn. This work is not yet complete, and further close examination of the MS can still contribute to it.
The two missing quires (numbers 16 and 18) coincide with gaps in the foliation (2 missing folios each). This tells us that the persons who added both sets of numbers must have been fully aware that there were other folios that are now missing. The most logical assumption is that they were not missing when the quires were numbered. The (later) foliator might just have skipped these numbers because he could have assumed that there was one missing bifolio in each case. However, quire 8 tells us otherwise.
Quire 8 is the first incomplete quire in the MS, and it consists of only two bifolios instead of the expected four. All seven preceding quires constist of the standard four bifolios. However, in this quire the folio numbers jump from 58 to 65, showing that there should have been three bifolios in the now missing part (59+64, 60+63 and 61+62), meaning that this quire would originally have had five bifolios. This cannot be the result of a guess.
The conservators who inspected the MS in November 2014 also clarified that the present binding must have been made while these three bifolios were already missing. It would have been impossible to remove them from the bound book without destabilising the complete binding. This clearly tells us two things:
The two types of herbal bifolios identified by Currier, one in hand 1 and language A, and the other in hand 2 and language B, seem to have become arbitrarily mixed. However, the quire numbers increase monotonously throughout this part of the MS, with the number always in the same location, regardless whether the bifolio is herbal-A or herbal-B. In particular, the folios 40 and 48, which have the quire numbers 5 and 6, are herbal-B folios, which all others are herbal-A. This strongly suggests that the quire numbers were added after these herbal folios got mixed up.
Quire 9 with its two foldout folios 67 and 68 was already mentioned. In the present layout, folio 67 is two panels wide, while folio 68 is three panels wide. The quire number (9) appears to be in the wrong place and John Grove already noticed that there appear to be stitching gaps of a previous binding in the foldout crease between the two panel of f67. By assumung that this should have been the original binding crease, all of a sudden the quire number falls in the right place, (see also Pelling (2006) (see note 6)).
On the other hand, while the quire number is now in the wrong place, the folio numbers 67 and 68 are in the right place for the present binding, yet would be in the wrong place if the binding had been made using the unused fold. In summary, the quire number is consistent with the presently unused stitching gaps while the folio numbers are consistent with the ones that are now used.
During the inspection of the MS in November 2014 I asked one of the conservators to have a closer look at these unused stitching gaps, and it turned out that these were only tiny cuts into the vellum made with a knife, i.e. preparing for a binding, but they were never used in any previous binding. Their location also does not match with the actual sewing stations.
One possible explanation is that this large foldout was originally prepared in one way, and the quire number written accordingly, but when it was to be integrated in the MS it was done in a different way, and the folio numbers were added at that time.
Finally, and curiously, the folio number 67 appears to have been erased, and then written again. It is possible that the erased number says 62 rather than 67. All in all, what really happened with this bifolio still requires further thought.
The main uncertainty about the production of the Voynich MS remains the question when the colours were added. The low quality of the painting has led some people to suggest that it was probably not done by the original composers, as it effectively deteriorated the quality of the work (7). It has also been suggested that the painting may have been done in different stages, by different people, e.g. one close to the completion of the MS and one much later (8). This is an important question because if the colours were added arbitrarily, they are misleading everyone who is trying to identify the plants.
More information about the order of painting can still be obtained by a more comprehensive analysis of the use of colours throughout the manuscript. This should include a complete overview of all the hues of the paints, and paint transfer between folios by contact. Until now, these aspects have only been looked at in piecemeal fashion, most actively by Nick Pelling (9), leading to results that are still very debatable. My present impression is that paint transfers are mainly of the dark blue paint, and they are consistent with the present binding, but this should be checked for the entire MS.
From an illustration on f33v, which disappears into the binding gutter and appears again on f40r, we can conclude that the painting was done before the (present) binding. Since the blue paint has also transferred onto the opposite pages, we can also say that this painting of blue pigment was done not long before the binding.
In addition, on folio 42 recto, the folio number has been written inside the drawing of a leaf. This leaf was also painted in green. In 2009, when the MS was inspected by microscope by Joe Barabe of McCrone, I asked him to look at this, in order to see whether the page number was written before or after the painting. The surprising conclusion was, that it really appeared as if the folio number was there before, and the paint (consisting of small green crystals) was over it.
The points that have been presented in relation to the order of production of the MS may now be summarised.
This leads to the following tentative reconstruction: