1918] Michael: Behavior of Salpa democvatica 261 



(1866) lias said, each aggregate salpa remains attached to its pre- 

 decessor after as well as before being pushed out of the mantle cavity 

 of the solitary salpa into the water. Does it seem unlikely, then, that 

 the chain also remains attached to its progenitor? 



4. Implications Regarding Locomotion 



If, as claimed on page 260, the general occurrence of protruding 

 chains is the only feasible explanation of all the temperature relations, 

 then some form of locomotion of the individuals of the two generations 

 is implied. For, to return to the corks and iron weights, the only 

 likely manner in which corks could be most abundant and least fre- 

 quent on the surface, while the attached iron weights were most 

 abundant and most frequent on the bottom is, as stated on page 257, 

 for corks with less than a given number (six) of attached weights to 

 float and for those with more attached to sink. That is, the presence 

 of weights on the surface would be consequent upon the buoyancy of 

 the attached cork, and the presence of corks on- the bottom upon the 

 sinking propensity of the attached weights. 



Translating back into Salpa terminology, the inference seems clear 

 that protruding chains of less than a given but unknown number of 

 aggregate salpae — short chains — are prevented from leaving an area 

 of warm surface water by virtue of the activity of the attached soli- 

 tary form, while with chains consisting of more than t^iis number — 

 long chains — the solitary form is prevented from escaping out of an 

 area of cold surface water by virtue of the combined activity of the 

 attached aggregate salpae. I do not mean to stipulate that aggregate 

 forms are negatively thermotactic and solitary forms positively ther- 

 motactic, nor that long chains move horizontally out of warm into 

 cold areas of surface water, while short chains move horizontally 

 out of cold into warm areas of surface water. This may or may not 

 be the true explanation, but it seems unlikely. Further, it is not 

 stipulated that aggregate forms are negatively geotactic and solitary 

 forms positively geotactic in cold surface water, while the responses 

 are reversed in warm surface water. This, again, may or may not be 

 the true explanation ; the data are inconclusive. Lastly, it is not 

 stipulated that an internal control by aggregate forms and solitary 

 forms over their own specific gravities leads to an accumulation of 

 short chains in warm surface water and long chains in cold surface 

 Avater ; nor is it stipulated that some form of metabolic rhythm is 



