1^1'^] Michael: Behavior of Salpa democratica 241 



The data involved in this investigation comprise all the collec- 

 tions made during the months of June and July prior to 1910 with 

 a surface net of OOOXX mesh silk bolting cloth having an orifice 97.5 

 cm. in diameter (Michael and McEwen, 1915). The reasons for 

 excluding later collections are first, that a net of very different 

 "catching capacity" has been in use, and second, that the species 

 has been captured too seldom and in too small numbers to permit 

 satisfactory statistical treatment. The reason for excluding data 

 relative to other months than June and July is made evident in the 

 preliminary discussion of the seasonal distribution of the species. 

 Lastly, while the vertical distribution of the species is briefly con- 

 sidered preliminary to the main discussion, too few subsurface col- 

 lections were made at any one depth corresponding in time and place 

 with the surface collections, to justify burdening this paper with a 

 consideration of details of little aid in interpreting the facts revealed 

 by the surface data. 



It is hoped that study of the distribution of this species may be 

 carried further in the near future. Does the species perform vertical 

 migrations? If so, to what depth does it migrate, and how is it 

 influenced by variations in light intensity, salinity, gas content, etc. ? 

 These are questions as yet unanswered. But sufficient attention has 

 been given them to make it evident that a more complex statistical 

 method than the one here employed is needed. Such a method has- 

 been devised, and is now being tested preparatory to publication. 



I desire to express my obligation to Professor W. E. Ritter, 

 director of this institution ; to Dr. F. B. Sumner, biologist ; and to 

 Dr. G. F. McEwen, hydrographer, for suggestions and aid in the 

 preparation of this paper. 



A. PRELIMINARY 



1. Description of the Life Cycle 



Salpa democratica, like other species of the genus, is notable for 

 its two alternating generations. One, known as the "solitary," 

 "budding," or incorrectly as the "asexual" generation (pi. 9, fig. 2) 

 produces offspring by budding, and the buds, when mature constitute 

 what is known as the "aggregate" or "sexual" generation (pi. 9, 

 fig. 3). Each individual of the aggregate generation then produces, 

 sexually, one of the solitary generation. So the cycle repeats itself. 



