318 
conditions between the summer floods of 1896 and those of 1897 
lies in the fact that in the latter year a great spring flood 
preceded the summer floods. This, it seems, might occasion 
the difference in summer production apparent on comparison 
of these two years. The slowly receding spring overflow of 
1897 seeded the submerged territory with cysts, spores, and rest- 
ing stages of the planktonts which afford the basis for rapid 
production upon the next flood invasion. In 1896 the summer 
floods follow a period of two years in which there had been no 
prolonged. overflow in a period of marked plankton production. 
The overflowed lands were thus not recently seeded, and pro- 
duction was longer in gaining headway in 1896 and did not at- 
tain the same amplitude. 
It is to be noted that this pulse arisesin declining nitrates 
and falls away as the temperature rises. 
The August pulse does not reveal itself plainly in the vol- 
umetric data, though it stands out. more clearly in the statis- 
tical curves (Pl. LII.). Adopting these as a clue to its limits, 
the pulse has a duration of 25 days—from the 30th of July to 
Aug. 24. The volumetric data would apparently terminate it 
on the 17th and limit it to 18 days. Its greatest amplitude is 
2.02 cm.’ per m.’*, attained on both the 3d and 10th. Accepting 
the shorter interval, the mean falls on the 10th, 21 days after 
that of the former pulse, while with the longer interval it is on 
the 14th—24 days after the preceding mean. This is a month 
of remarkably uniform production, the departure from the 
mean in no case exceeding 26 per cent. It is accompanied by 
stable hydrographic conditions, the total movement being 2.2 
ft., most of which occurred in the first week. After a heat 
pulse in the first week the temperature conditions were also 
stable, while in the chemical conditions there is but slight 
change. The chlorine increases as the sewage contamination 
rises with decline in levels. The absence of any marked max- 
imum in this month is evidently due to the fact that the ani- 
mal plankton which forms the greater part of the volumetric 
pulses has not greatly fluctuated as a whole. The occasion for 
