1918] Michael: Bcliavior of Salpa dcmocratica 271 



It seems, then, that whatever the specific gravities of individuals 

 of either generation may be, and whatever may have been the way in 

 which viscocity of the water was related to high and low temperatures, 

 the observed differentials could not have been caused by changes in 

 viscocitj' alone. Furthermore, as any combination of physical influ- 

 ences could, at best, only directly affect one generation more than 

 the other, it follows that whenever the generation least liable to be 

 affected was affected, the other generation must of necessity have been 

 similarly affected. This means that the observed differentials could 

 not have thus arisen. Is it not evident then, that, although changes 

 in viscocity, etc., have doubtless affected the magnitudes of the 

 observed differences in abundance of the two generations, such changes 

 alone are impotent to explain the differentials noted 1 



D. THEORY OF LOCOMOTION 

 1. Behavior of Chains 



In the preceding pages it has been demonstrated that neither differ- 

 ences in season, differences in position, nor a combination of these, 

 nor variations in the condition of the water are adequate to account 

 for the facts that : (1) solitary forms are most abundant on the surface 

 when the temperature of the water is high; (2) aggregate forms are 

 most abundant on the surface when the temperature is low; and (3) 

 both solitary forms and aggregate forms are most frequent on the 

 surface when the temperature is low. Furthermore, tables 5 and 6 

 make it evident that influences associated with time of day could not 

 have given rise to these relations. Is the conclusion, then, not forced 

 upon one that, as stated on page 262, "the only type of behavior con- 

 sistent with all the facts is some form of locomotion"? Since the com- 

 bined effects of all surface influences directly associated with time and 

 space, i.e., with time of day, day of the month, month of the year, 

 and with latitude and longitude, do not materially alter the observed 

 differentials, locomotion appears to be the only instrumentality by 

 means of which they could have arisen. 



It is characteristic of all salpae, as stated on page 242, that when- 

 ever they breathe or feed they move forward along a stream of water 

 forced through their bodies. Concerning the movements of this 

 species (described as S. cabotti) Agassiz (1866, p. 18) says: "The 



