337 
significant that the nitrates fall during this period of increased 
production (Table X.) to half the content in November. The 
average production in this month (.99) is in excess of that in all 
previous years excepting 1895, when stable conditions of longer 
prior duration were prevalent. The relation of stability in en- 
vironmental conditions to increase in plankton production is 
thus confirmed by the data of this month. 
As a whole, 1898 was a yearof relatively light plankton 
production, averaging 2.13 cm.’ per m.’ (mean of all collections), 
or 2.03 (mean of monthly averages). This is all the more 
apparent when we note (PI. XII.) that production falls below 2 
em.’ in all but eight weeks of the year and below | in all but 
fourteen. The only large production is found in April—June, 
and the unusual extent of this brings up the yearly average. 
The well-defined vernal pulse under peculiarly favorable hydro- 
graphic conditions, and the suppression of production by the 
flushing effect of repeated floods are the prominent features of 
the year’s record. The effect of stability of environmental 
factors in increasing production, and of instability in suppress- 
ing it isapparent. The cyclic movement of production is also 
to be traced throughout the year. 
1899. 
(Tables III., X.; Pl. XIIL, XLV., LIL.) 
Collections at weekly intervals were made in this year 
through the month of March. The 15 collections afford an op- 
portunity of tracing the effect of the interrupted ice blockade 
which continued during the first two months in semi-flood con- 
ditions, and of noting the effect of the early maximum spring 
rise upon production. 
The January pulse has a duration of 21 days,—from the 10th 
to the 3lst,— with a maximum amplitude of .5 cm.’ per m.‘ on 
the 17th. Its mean falls on the 18th, 27 days after that of the 
preceding pulse. This is a month of gradual though consider- 
able change in levels, the total movement being 3.5 ft. The 
gradual character of the rise permits the development of a 
