262 U)iivcrsity of California Publications in Zoologij [Vol. IS 



involved. Once, again, either of these may or may not be the true 

 explanation, but neither seems likely. What the unescapable implica- 

 tions of the data are, may be listed as follows : 



1. Short chains accumulate in warm surface water and long chains 

 in cold surface water. 



2. Since the only difference between short and long chains con- 

 sists in the number and size of the attached salpae, a differential in the 

 distribution of the two generations must be due to a differential in 

 behavior of short and long chains, which obviously implies a differen- 

 tial in behavior of the solitary form and aggregate forms constituting 

 the chains. 



3. The only type of behavior consistent with all the facts is some 

 form of locomotion. 



However, these are implications — not facts — and it is necessary to 

 gain some idea of their reliability by reexamining the data. 



C. REEXAMINATION OF THE DATA 



1. Relation Between Season and Temperature 



It has been suggested that the difference between short and long 

 chains may actually not have been restricted to the number and size 

 of the attached salpae. This, it is said, would be true only if all 

 possibility of seasonal influence were eliminated. In other words, 

 it is argued that, owing to an intimate association between increasing 

 temperature and advancing season, even within the limits of June and 

 July, 1908 and 1909, the relation between temperature and abundance 

 of the two generations may have been consequent upon an increased 

 production of solitary forms and an increased death rate of aggregate 

 forms. 



On first thought this seems reasonable, but careful consideration 

 proves it untenable. For each individual of the aggregate generation 

 can give birth to only one solitary form, so that an increased produc- 

 tion of the latter requires an increased number of the former. Further- 

 more, as stated on page 245, the embryo is carried and developed within 

 the body of the aggregate form and is not set free until after it has 

 reached maturity and its stolon has begun to be eonvei-ted into the 

 salpae of the succeeding aggregate generation. Consequently, all 

 hauls obtaining an excess of solitary forms over aggregate forms would, 



