78 University of California Publications in Zoology. [Vol.6 



that the number on any particular fold in any particular animal 

 increases with the size of the animal. 



As table II shows, the average number of branchial tentacles 

 in the first thirteen, i.e., the thirteen largest individuals, is 46.6 

 while the average for the second twelve is only 38. Because of 

 the difficulty of counting these structures the difference in the 

 averages here is not large enough to make it sure that the results 

 indicate what they seem to. The probability is however that 

 they do. 



Does this numerical increase of parts with size of animal 

 mean increase with age? In other words, can we rely on size in 

 tins animal as an index to age? Of course this must be so with- 

 in certain limits. Stated exactly the question is: Given two 

 specimens, the one larger with a greater number of dorsal lan- 

 guets, for instance, does it follow that this one is older :' The 

 answer probably is that generally, though not invariably, this is 

 true. The data are insufficient to enable us to speak more posi- 

 tively. Prom the taxonomic standpoint however, it seems that 

 somewhat more account should be taken, in some groups of ascid- 

 ians at least, than generally is of the possibility that numerical 

 variations indicate age rather than specific difference. In some 

 genera, as for example Perophera, it is certain that a limit of 

 individual growth and a maximum number not only of stigmatic 

 series, but also, within rather narrow bounds, of individual stig- 

 mata, is early reached. 



5. To WHAT EXTENT MIGHT THE DEVELOPMENT OF THIS ANIMAL 

 BE REDUCED TO A REPETITIVE BASIS AND TREATED NUMER- 

 ICALLY .' 



Ascidiologists do undoubtedly take it for granted that, within 

 the same species at least, difference in size, and with this, differ- 

 ence in number of repetitive parts means difference in age. This 

 assumption surely contains the truth, but the truth only in a 

 general way, we say. Now the question comes, as it is hound to 

 in every branch of science, how long are we justified in being 

 content with procedure and hence with results based on what is 

 true "in a general way"? Sooner or later any science, if it is 

 really going ahead, is sure to look more closely at assumptions 



