314 
curve to the left, giving a left-handed skewness to the curve. 
Aside from this depression due to flood there is a general de- 
cline in production as levels fall, the pulse closing on the 21st 
with a minimum of .28¢m.’ This decline in production is at- 
tended by a steady rise in nitrates, organic nitrogen, and free 
ammonia (Pl. XLIII.), and thus in the presence of increasing 
nutriment, as well as growing hydrographic stability—that is 
lower river levels. The two summitsof production in this pulse 
coincide with temperature pulses. The plankton and tempera- 
ture pulses are alike set off by flood waters, and the causal 
nexus may lie between plankton and flood rather than between 
plankton and temperature. The location of the flood also has 
the effect of lowermg the average production of the month to 
1.12 cm.*—the lowest average on record, excepting only 1898 
(.91), also a year of much disturbance. 
From this point tlte remaining collections of 1896 are too 
infrequent to delineate or even to suggest recurrent volumetric 
pulses. They are also insufficient to adequately trace the results 
of hydrographic changes. Diminished production at the time 
of rapid decline of temperature is apparent late in September. 
Increased production follows declining flood and stable tem- 
peratures and a downward movement of nitrates (Pl. XLIII.) 
in October, and the phenomenon is apparently again repeated 
in December, though the volumes do not equal those of the 
preceding year. 
As a whole, 1896 was a year of but slight plankton produc- 
tion, averaging only 1.16 (average of all catches) or 1.05 (aver- 
age of monthly averages) cm.* perm.’ This is only a half or 
a third that of other years in our records (see table on p. 292). 
The silt, on the other hand, is more abundant than in any other 
year, averaging 2.55 cm.’ per m.’ to .28, .72, 1.91, and 2.11 re- 
spectively for 1894, 1895, 1897, and 1898. The total movement 
in levels for this year at Copperas Creek is 53.16 ft., an excess of 
that in all other years but 1898 of 4 to over 40 per cent. (Ta- 
ble I.). From this fact,and from the evidence accumulated in 
the detailed discussion, it is apparent that the oft-recurrent 
