78 Universitij of California Puhlications in Zoology. [Vol.8 



greater, the following method was employed. After sorting the 

 species, from sixty to one hundred specimens were removed 

 with forceps and placed on filter paper to allow the fluid to 

 drain off. This sample was then accurately weighed and 

 counted. Dividing this weight by the number of specimens in 

 the sample gave an average individual weight. A second sample 

 was then treated likewise. The mean between the two average 

 weights thus obtained was considered final, and the weight of 

 the entire haul divided by this mean approximated the number 

 of individuals contained. The usual error in this method has 

 been calculated from twenty-five tests and amounts to about 5 

 per cent. In no instance was it over 10 or under 2 per cent. 



