Bifolia created using previous as reference

[npcompl33t]

Here are the non descriptive results. I looked for chains that appear to exist between the bifolia, and compared them to the existing Golf 2.1 structure.

Here is an example of how the statistics are compared. 

   

Here are the overall chain sequences that I identified. 

   

The sequence at the beginning of the MS contains the largest chain.  Its possible that this layout method was not used consistently across the MS, but there appear to be strong signals across Herbal A, Stars, and Bio. Herbal B contains far fewer obvious matches. 

Additionally, several 'changes' to the current golf ordering are considered. These mostly involve ambiguous sequences within the golf ordering, that the new evidence provides additional insight into. 

For example, the current golf ordering has D:27:30 as occuring before A1-8. This analysis suggests these should be swapped. 


The chart below shows the changes, along with the actual statistics in the table below. 

   


This results in the following full chain. 

   


The other proposed golf swaps are as follows. If you are interested in seeing the same plots as above for these, i can provide them, but it will quickly become a lot of images so I didn't add them all. 


The changes are: 

• Herbal A / 1a: A1-8 before D27-30 (shown above)
  
• Herbal A / 1a: A:3-6 -> C:17-24 -> G:52-53 becomes G:52-53 -> A:3-6 -> C:17-24
  
• Herbal B/ F58: D:26-31 -> H:57-66  this one is interesting, and may suggest one herbal B page actually was written between Ff8 and f57
  
• Heral B: F:43-46-> F:41-48 (current), these are swapped to F:43-46-> F:41-48 (new)
  
• Im double checking the Bio pages against Layfied and Davis two reported orderings, once i have those tests i will share them.