206 



This was not true in other years, but collections in those years were 

 too infrequent to trace the seasonal distribution of the species with 

 accuracy at that season. It is volumetrically of some importance 

 in determining the quantitative fluctuations in the total plankton. 

 Computations based on its average size indicate that approximately 

 600,000, including eggs, would be required to form 1 cm. 3 of plank- 

 ton. On this basis, and allowing 10 per cent, for interstices, it 

 constituted at the time of its vernal maximum in 1898 about 10 

 per cent, of the total volume of the plankton (silk-net catch). 



The table on pages 204 and 205 lists 43 pulses, of which 6 lie out- 

 side of the period included in Plates I. and II. Of the 38 remaining 

 pulses' 16 coincide in location with the whole or a part (in case of 

 divided culminations) of the pulses of the chlorophyll-bearing organ- 

 isms; 12 follow at the next collection, usually at intervals of one 

 week ; and 6, after a fortnight. The remaining 4 do not bear this rela- 

 tion, occurring in autumn or midwinter, when all pulses were feeble 

 and ill-defined. A comparison of Plates I. and II. with V. will 

 show that not all of the chlorophyll-bearing pulses are attended by 

 pulses of Polyarthra; nor is there any constant relation, excepting 

 the vernal pulse, between the size of the pulses of the two groups 

 of planktonts in question. Nevertheless, the dependence of the 

 recurrent periods of rapid multiplication of Polyarthra upon the 

 rhythmic occurrences of the chlorophyll-bearing organisms upon 

 which they largely depend for their food is strongly suggested by 

 the data here offered. Food relations thus dominate the repro- 

 ductive cycles. 



The pulses of Polyarthra form a considerable portion of many of 

 the pulses of the total Ploima, and it is but natural that we should 

 find a coincidence in their locations. This may be followed for 1898 

 in Table I . In a number of instances the culminations of the pulses 

 are not exactly coincident, but separated by the interval between 

 two collections. The association of the two pulses is, however, 

 apparent in every case, and a similar relation may be traced in prior 

 years. 



These recurrent pulses afford evidence for the polycyclic habit 

 of this species. Additional proof of this phenomenon is found in the 

 evidences of sexual reproduction either male or winter eggs 

 attached to the female which have attended many of the pulses. 

 The eggs of this species, both summer and winter forms, are very 



