VdL. XVI 

 1893 



] PROCEEDINGS OF THE NATIONAL MUSEUM. 117 



SO that the whok^ presents the aspect of a very fine ccJhihir tissue sim- 

 ihir to that which is found occupying the intercellular s])aces of trans- 

 verse sections. Into these open areas the large tubular cells are found 

 to project in a vermicular manner, precisely as in perfectly preserved 

 specimens of Nemntophjiton Lof/ani and other species exanuned by me. 

 The tubular cells are in no case perfect, but sufliciently so to indicate 

 their original character. No evidence of intercellular lilaments could 

 be found. 



Comparing these specimens with the ty\w of I^oiKdophyton crassnm,* 

 we find they agree with it in all respects except the absence of inter- 

 cellular filaments from the former and their i)resenc(^ in tlu^ latter. 

 But this difference may safely l)e attributed to the opeiation of greater 

 alteration in one case than in the other, and it is therefore admissible 

 to consider that my reference of CelluloxyUm primwvum to NcnKdophytou 

 cras-suiu was not only correct, but that it receives striking conliiination 

 from these specimens. 



It may also be well to ])lace on record a few observations made dur- 

 ing my examination ol' this material, as bearing upon the alteration of 

 organic; structure by decay and crystallization. 



The extent of alteration ai)pears to dejx'iid in tlie lirst instance upon 

 the extent of decay in the organic stincture at the time when crystal- 

 lizaticm of the infiltrated silica becomes pronounced, and thus upon 

 the conditions favorable or adverse to freedom of growth in the crys- 

 tals. This is clearly shown by the transitional forms of the structure 

 as already described, which, in turn, also show tliat the imi)erfect tub- 

 ular structure seen in longitudinal section and the large parenchyma 

 like cells of the typical Gelluloxylon are derived, not from the tubular 

 cells of the original structure, but from the spaces surrounding and 

 lying between them; that is to say, ciystals or groups of crystals form 

 in the intt;rcellular si)aces and, tinally, in the cell cavities in such a 

 way as to crush the tubular cells into shapeless masses of carbon, 

 which afterwards become mor(i or less broken up or remain as large 

 and irregular masses of carbon at the angles of the Celluloxylou cells. 



Three stages in the (;onversion of the normal structure may be noted: 



(a) Conversion of the intercellular hyplne, the medullary stru(;ture 

 remaining largely intact. This results in the fin-mation throughout 

 the intercellular spaces and in the ojx'u tracts of a fine (k'Hidoxylon 

 structure, due to the aggregation of numerous small crystals of silica, 

 upon the surfaces of whi(;h the carbonaceous products of decay are 

 deposited. This gives to the present specimen tlie peculiarities of 

 structure which distinguish it from tiie' typical A^. crassran. 



[b) Conversion of the intercellular hyplne and i)artial conversion of 

 the medullary structure the tubular character of whi(;h is nevertheless 

 evident. There is also in this condition a i)artial formation of the 



*Traus. Royal Soc. Can., vii, iv, 25, PI. i. Fig. 5. 



