24 PENHALLOW ON" 



noted a variable tendency towards the redistribution of the organic matter, with a 

 consequent breaking vip of the normal strnctvire and its replacement by a granulated, 

 carbonaceous substance which always tends to be determined along the lines of contact 

 between contiguoi^s crystals. It is obvious, therefore, that if the proper conditions of 

 structure are present in the first instance, together with sufficient growth in the crystals, 

 the organic matter will not only suffer complete redistribution, but at the same time will 

 take up such positions as to produce a false cellular structure which will have the same 

 general character in both longitudinal and transverse sections ; although, in the former, 

 there will always be a tendency for the false cells to be distributed in lines, as determined 

 by the direction of the cells from which they were derived. This view, we find, was 

 advanced by Sir Wm. Dawson in one of his earliest papers on Prototaxites, when he 

 pointed out that, " In parts of the larger trunks, as is usual with fossil woods, it has been 

 replaced by a concretionary structure, or by that pseudo-cellular structure which pro- 

 ceeds from the formation of granular crystals of silica in the midst of the tissues. 



" In fossils woods the carbonaceous matter, being reduced to a pulpy mass, some- 

 times partly becomes moulded on the surfaces of hexagonal or granular crystals in such a 

 manner as to deceive, very readily, an observer not aware of this circumstance." ' 



The statements thus made admit of ready confirmation, since it is precisely what 

 occurs in many specimens of Nematophyton, and not only do we find plants in which the 

 entire structure is thus transformed, but in others, where the normal structure is pre- 

 served, there are often found localized tracts within which siich alteration has taken 

 place. From facts brought to my notice, I am led to consider a more or less advanced 

 condition of de(^ay as preceding the impregnation by silica, and thus essential to the 

 changes noted. 



These statements are partly based upon the following observations, which will serve 

 to indicate the correctness of my position. 



No. 1. — A transverse section. The organic matter was found to be nearly eliminated, 

 but still enough was present to show it sredistribution to such an extent as to completely 

 destroy the normal structure and give rise to a false cellular structure, closely resembling 

 that of Celluloxylon in all its essential features. Upon measurement it was foiind that ten 

 cells, selected at random, gave a mean diameter of 60 /< and an extreme variation of from 

 40 /' to 90 ).i. 



No. 2. — A silicified specimen in which the organic matter was much more abundant, 

 but subjected to precisely the same redistribution and closely resembling Celluloxylon. 

 Ten measurements gave an average diameter of 62 /< and an extreme variation from 40 /< 

 to 130 1.1. 



No. 3. — A sicilified specimen in which the tubular cells were occasionally found, but 

 for the most part the organic matter was redistributed as in the previoiis cases. Ten 

 measurements gave an average diameter of 61 /< and an extreme variation of from 40 /< to 

 90 //. In this specimen it was also noted that still larger cells were obviously formed by 

 coalescence of two or more smaller cells, and this fact explains the unusually large one 

 observed in No. 2. 



No. 4. — A silicified specimen in which the normal structure was largely preserved, 



' Montli. Mir. ,7n'l., x. G9, 70. 



