298 



of the initial stages of the rise in vernal production indicated 

 in later years, especially in 1896 and 1898. 



The two April collections average 3.18 cm. 3 — about 37 per 

 cent, below the mean for this month in the three subsequent 

 years. An inspection of the vernal plankton curves for this 

 and subsequent years at this and other stations (PL IX.-XIL, 

 XXVI.-XXXIV., XXXVI.-XLII.) reveals the presence of a ver- 

 nal volumetric pulse* of plankton, which, as a rule, marks the 

 maximum period of production in the year, and follows imme- 

 diately upon the vernal rise in temperature. 



In 1895 our collections were too infrequent to detect the 

 location and extent of this vernal pulse. In other years, as 

 seen in the plates to which reference is made above, the best- 

 defined vernal pulses appear in the closing days of April and 

 the first week of May. From the character of the best delineat- 

 ed vernal pulses — e. g. those of 1896 and 1898 (PL X. and XII.) 

 — it is probable that the apex or crest of the pulse is narrow, 

 that is, the maximum development lasts but a few days. If 

 this be the case, our two collections in April may miss entirely 

 the period of culmination. The second collection, upon the 

 29th, would appear to be located at the probable season (tem- 

 perature?) to detect the maximum development. If this be the 

 case the vernal development of 1895 is much reduced, and might 

 be correlated with the suppression of overflow stages and con- 

 sequent reduction of contributions from the impounded back- 

 waters. Two facts lead me to think that two well-developed 

 vernal maxima may have been present in 1895. First, a com- 

 parison of the vernal pulses of 1896 (PL X.) and 1898 (PL XII.) 

 indicates that the pulse of the former year culminates about 



*I use the term plankton "pulse" to designate the phenomenon of a periodic in- 

 crease of the plankton volumetrically, as a whole, from a minimum to a maximum, 

 followed by a decline to another minimum, the rise and fall being more or less 

 gradual, and the data forming when plotted a more or less symmetrical curve, re- 

 sembling that known as the "probability of error" curve. A typical example of this 

 phenomenon and resulting curve is seen in the case of the April-May plankton of 

 1898 (PL XII.). 



The term may also be applied to a periodic increase in individual members of the 

 plankton similar in its graphic delineation to that of volumetric changes. The 

 pulses will be designated by the months in which the major part of their course is run. 



