237 
May 25 in 1897, at 66.3°; and on May 24, in 1895, at 73°. From this 
maximum the pulse declines abruptly in a fortnight to a midsummer 
minimum during maximum temperatures, which continues until 
September. During this period the numbers are small, rarely rising 
above 400 per m.° (average, 388, 103, and 49), and the occurrences 
are also less numerous (72, 61, and 33 per cent.). With the decline 
of temperatures which begins in September the percentage of occur- 
rences mounts to 81, and the average per m.* to 228, and remains 
near this level during the remainder of the year. 
An analysis of the full statistical data, of which the records for 
1898 are fairly typical, confirms the conclusions drawn from these 
averages. Chydorus in channel waters is monocyclic, with a well- 
defined vernal pulse in March—June which includes 95 per cent. of 
the total annual Chydorus population. There are suggestions of an 
autumnal pulse, but the data are not sufficient to delimit it. There 
is no satisfactory evidence that there are recurrent cycles or pulses 
at briefer intervals during the year. 
The dominating effect of temperature as a regulating factor in 
delimiting the seasonal distribution of Chydorus is very evident. 
This, inaddition to its appearance in the annual curve ofoccurrences, 
is also exhibited most clearly in a comparison of the vernal pulses 
in the two years of fullest representation in our records, 1896 and 
1898. The following table gives the data of dates, temperatures 
of surface waters, and numbers of Chydorus. 
From these facts it appears that the late spring of 1898 delayed 
the vernal pulse of Chydorus, and that the early spring of 1896 
accelerated it in that year so that their apices (April 29 and May 24) 
are four weeks removed from each other in seasonal location. In 
both years the rapid rise in the pulse appears after 60° 1s passed, the 
culmination occurs at about 70°, and the decline, 7 temperatures 
above 70°. 
Egg-bearing females were more abundant during the rise of the 
pulse, and less numerous during its decline. Evidence of great 
mortality during the decline of the pulses is to be found in great 
increase in the relative numbers of empty carapaces. Thus, during 
the decline of the vernal pulse in 1896 there were on the day of 
culmination, April 29, 2,780 dead to 18,904 living, on May 1, 3,570 
to 14,875, and on May 8, 1,578 to 6,706. From 14 to 24 per cent. of 
the Chydorus population had thus recently perished. Parasitized 
