360 



PROCEEDINGS OF THE NATIONAL MUSEUM 



than the March or November generations. If this explanation is 

 correct, it means that several generations are produced each year, 

 with each generation living only a few months or less. In contrast, 

 Dunbar (1957) proposes a life cycle of 18 months to 2 years for P. 

 libellula in the eastern Canadian Arctic, and Bogorov (1940) gives evi- 

 dence for a 2-year life cycle for P. abyssorum in the Barents Sea. 



75r- 



70- 



6.5- 



* 60 



k. 



>- 5.0 

 4.5 



4.0 



13 



14 



15 



16 



17 



T EMPE RATURE ( °C ) at 30 M. 



Figure 10.— Parathemisto pacifica Stebbing, relation between body length and tempera- 

 ture at a depth of 30 meters. Plus signs, cruise 1, March 1949; circles, cruise 5, July, 

 1949; triangles, cruise 9, November 1949. Each symbol represents the average length 

 ol the adult females from one station. 



However, there is considerable evidence, much of it summarized by 

 Dunbar, that marine poikilotherms from warm water grow to maturity 

 more rapidly than do those from cold water. It is reasonable to 

 suppose that P. pacifica passes through its life cycle more quickly 

 than its Arctic relatives. 



2. The Parathemisto population sampled in July may be older, 

 hence larger individuals of the generation sampled in March. The 

 November population, composed of smaller individuals than the July 

 population, must represent a different generation. This interpretation 

 requires that an individual breed more than once during its lifetime. 

 Dunbar concluded that P. libellula breeds only once and then dies, but 



