1910-11.] Mathematical Theory of Random Migration. 289 
(3) It is found in both cases that the exponential curve can be taken as 
giving a fair representation of most of the facts. The manner in which 
this curve, however, appears is somewhat different in the two instances. 
It is doubtful why this formula should express both results. 
Note. — The experimental work recorded in this paper was done chiefly 
at the laboratory of the Millport Marine Biological Station, and I desire 
to thank the Superintendent for much assistance. 
(. Issued separately February 25, 1911.) 
£XXL 
19- 
