1^18] Michael: Behavior of Salpa deniocratica 253 



1. Solitary forms most abundant in the warmer water. 



2. Solitary forms most frequent in the colder water. 



3. Aggregate forms most abundant in the colder water. 



4. Aggregate forms most frequent in the colder water. 



5. Frequency of solitary forms very similar to that of aggregate 

 forms. 



6. Abundance of solitary forms reversed to that of aggregate 

 forms. 



With respect to aggregate forms these relations seem reasonable, 

 for the orders of their abundance and frequency are parallel as is to 

 be expected. Regarding solitary forms, however, the relations appear 

 to be meaningless. Do they not signify that the solitary forms are 

 found in greatest abundance in the places they frequent the least or, 

 to state it differently, that they are found most often under the con- 

 ditions when they occur in smallest numbers? Obviously, this is 

 precisely what is implied if the relations revealed by table 4 are not 

 due : ( 1 ) to systematic errors introduced by the method of collecting ; 

 (2) to the chance effect of random sampling; or (3) to what amounts 

 to the same thing, an insufficient number of hauls. 



First, as to methods used in collecting. All collections were made 

 with surface nets of 000 mesh (Michael and McEwen, 1915, p. 201), 

 as nearly like one another as it is possible to construct them. Further- 

 more, the "Agassiz" was allowed to drift during every haul, and the 

 error introduced by variation in volume of water filtered is far within 

 the differences in abundance noted in table 4. The ratio between the 

 mean variability (0.172 km. per hour) and average velocity of tow 

 (1.8 km. per hour) as determined from fifty hauls made with a cur- 

 rent meter attached below the net, is 0.01, while the mean variability 

 in number of organisms obtained per hour usually exceeds the average 

 number (Michael, 1916, p. xviii). It is unlikely, therefore, that lack 

 of standardization of nets or method of collecting could have been 

 responsible for the relations above revealed. 



It may be urged, however, that the effect of random sampling, i.e., 

 an uneven distribution of hauls with respect to influences other than 

 temperature, might account for the above relations. This is partly 

 true ; the hauls were unevenly distributed with respect to light, only 

 nine night hauls (6 p.m. to 6 a.m.) having been made in the colder 

 water, while thirty-three were made in the warmer water. The follow- 

 ing table, however, includes only night hauls and the same relatians 

 persist. 



