L909] Hitter: Halocynthia johnsoni a. sp. 77 



in number of the languets with increase in size of the indi- 

 vidual is undoubted, and the average number for the thirteen 

 largest, 113, as compared with that of the eleven smallest, 74.4. 

 is too pronounced to admil of any other interpretation than thai 

 the number of dorsal languets increases with increasi in sizt of 

 tin a iii mills. The data are no1 sufficient to prove whether the 

 increase continues as long as there is increase of size of the indi- 

 viduals. So far however as the evidence goes it indicates such to 



lie the case. 



Xo effort was made to bring the change of shape of the stig- 

 mata to a numerical basis, though beyond the labor of making 

 the measurements it would not be difficult to do. The change 

 referred to consists in an increase of the ratio of length to breadth 

 of the orifices. The classification into "lone" and "short" can 

 of course present the actual conditions only in a rough way. 

 Nevertheless when all the branchial sacs are examined somewhat 

 attentively by the same person (Miss .Johnson did this', the line 

 of separation between "lone" and "short" is sure to In- no1 far 

 from the mean as this would be obtained by measurement. That 

 the result reached was not subject to the bias which previous 

 knowledge or supposition that length of stigmata and size of 

 animal go together, is assured from the circumstance that this 

 possibility was not thought of by .Miss Johnson until the original 

 laboratory notes had been tabulated. As a matter of fact the 

 classification was made with the supposition of quite a different 

 correlation. Figures 15, 1<>. and 17 illustrate the increase in 

 length of stigmata with increase in size of animal, the individual 

 from which fig. 15 was drawn being 1.25 cm. long, and that from 

 which fie. Iii was drawn being •"> cm. long. The evidence that the 

 longitudinal vessels of the branchial folds increase in number 

 with size is Utile if any less convincing than is that of increase 

 in number of dorsal languets, and in proportional length of stig- 

 mata. 



The average for each individual is found by adding together 

 all the vessels of all the folds of one individual and dividing by 

 the number of folds. This does not regard Hie number on any 

 particular fold, i.e., the distribution of the vessels on the folds. 

 I have little doubt that were the data sufficient it would be found 



