GRAPTOLITES OF NEW YORK, PART 2 



79 



from the other spines as in the first named species, or they alone may 

 bear spines, and that these spines may grow to such extreme size that the 

 thecae are thickened into spine bases [see text fig. 6 J similarly as those 

 of Goniograptus thureaui are into stems, losing thereby their 

 normal functions. It is in these cases obvious that the first two thecae 

 have, by division of labor, been assigned to the task of protecting the most 

 exposed part of the colony. In some mutations the virgella or sicular spine 

 also grows out to great size and the sicula shares in the defensive work. 



Also the distribution of the spines on the first six or eight thecae of 

 the rhabdosome in Dicellograptus while, as stated before, originally due to 

 the exposure of these thecae to the strongest external stimuli, is in some 

 degree indicative of a differentiation of the thecae. 



The appearance of the spines on the first thecae, which has been just 

 discussed, seems to contradict the former statement of the writer that " the 

 rhabdosomes in toto and their parts, the branches, seem to pass through 

 stages which suggest phylogenetically preceding forms," for the spines 

 which as new features ought to be expected first on the last thecae or the 

 astogerontic part of the colony appear in the first thecae of the colony. 

 But it must not be forgotten that these spines are not so much an 

 expression of waning vital power and originate intrinsically, as the result of 

 external stimuli at the places where the impact of the latter is most fre- 

 quently felt. They are hence rather of the character of new acquisitions 

 by external physical conditions than of internal physiological changes and 

 in this regard only in a general way indicative of the paracmic condition of 

 the class but not of their special phylum. 



Similar observations could be made in regard to the spinose character 

 of the sicula. There are forms where the sicula alone is spinose. The 

 most notable of these are Cryptograptus antennarius and 

 Cryptograptus tricornis. In the latter the sicula is even fur- 

 nished with four long spines [see text fig. 7]. From this it can be con- 

 cluded that the distal part of the sicula, which is provided with growth 

 lines, is not to be considered as an embryo sheath, for the embryonic sub- 



