I'^lS] Michael: Behavior of Salpa deinocratlca 259 



more or less advanced and resemble in every respect, long before it 

 becomes detached, the chains which are found floating about. These 

 sections are thus liberated in turn, new ones continually forming at 

 the base of the geminiferous tube during the budding season. . . . 

 These chains escape through an opening formed at the proper time 

 through the tunic [test] , near the nucleus on the ventral side, which 

 shows afterwards no trace of the passage of the small chain. ' ' Agassiz 

 repeatedly refers to these liberated chains so that, if jS*. cabotti is in 

 reality only a variety of S. democratica, it is evident: (1) that the 

 chain of aggregate forms remains intact after, as well as before, escap- 

 ing from the mantle cavity of the solitary form ; and (2) that the chain 

 is liberated in blocks of from forty to sixty individuals. Obviously, 

 this means that protruding chains must exist during the period of 

 liberation, but what length of time this involves is not indicated. 



Aside from these three instances I have failed to find a single 

 statement applicable in any way to the question as to whether or not 

 the salpae remain attached in the form of a protruding chain. In 

 fact the attitude of those familiar with the group seems to be either 

 that such chains may be assumed to exist on general principles, or 

 that they do not exist in this species. 



But negative evidence is never conclusive. Following up the 

 implications of tables 4 to 6, examination of the crude data (table 1) 

 shows that every haul that captured more than six solitary forms, 

 and all but three capturing less than six, also captured aggregate 

 forms. The three that failed (hauls 1650, 1653, and 1804) were all 

 made in water exceeding 18?9C. Moreover, as aggregate forms accu- 

 mulate in cold, and solitary forms in warm surface water, a con- 

 siderable number of hauls made in the colder water ought to have 

 captured aggregate forms, but no solitary forms, if protruding chains 

 or salpae containing mature embryoes were not frequently encount- 

 ered ; but not a single such haul was made. Again, hauls made in the 

 warmer w^ater ought frequently to have captured solitary forms, with- 

 out aggregate forms, if protruding chains were not generally encount- 

 ered ; but only three such hauls were made. 



Further examination of the crude data reveals three more hauls 

 (1860, 1872, and 1875) made in water exceeding 18?9C which cap- 

 tured 21, 12, and 89 aggregate forms respectively, but no solitary 

 forms. Why? In answering this question it must be recognized that 

 during the processes of washing the net and of condensing, separating, 

 and examining the hauls a few specimens are nearly always lost. Con- 



