Occasional Papers of the Museum of Zoology 9 



mode would presumably transform before the next spawning 

 season. 



Summarizing the data contained in the curve, it is found 



that the annual groups may be arranged around the modes as 



follows: 



Age Thunder Bay River Gilchrist Creek 



3 months 20 millimeters 20 millimeters 



iji years 47.5 " 50 



2% " 60 " 65 



3}i " 70 " 75 



aVa " 80 " 85 



5J4 " 90 " 95 



6% " 105 " no " 



At present there is not a sufficiently large collection of lar- 

 vae from the type locality (Mill Creek, near Dexter, Michi- 

 gan) on hand for the study of the life cycle. It is hoped that 

 such a collection may be obtained so that a comparison may 

 be made with the above more northern series. It is possible 

 that temperature may have some influence on the length of 

 the life cycle and that the northern form may pass through its 

 development more slowly than the more southern forms. The 

 difference in latitude between the two points is only about 

 185 miles, but the difiference in, the temperature of the water 

 is greater than indicated by the difference in latitude. 



Rate of Growth. — Taking the average size of the larvae of 

 each year to be represented by the modes of the graphs i and 

 2, growth and growth increment curves may be constructed 

 which will throw interesting light upon the rate of growth in 

 the larvae during the successive years of their life cycle. Such 

 curves are shown in Fig. 3. 



It will be seen that the greatest increase in length takes 

 place during the first year and a quarter of the animal's life. 

 In fact, the larvae have attained to about half the adult length 



