— 72 — 



oscillations, so that we may assume tliat the water-level in thèse 

 raonths remained fairly uniform and that no very sudden altér- 

 ations in the concentration of the water took place. In 1006 

 on the other hand the month of April (i. e. a very critical time) 

 had an exceptionally low rainfall and a large amount of bright 

 sunshine, and as we know (cf. footnote 3 on p. 61) this led to a 

 marked shrinkage of the pond, involving a relatively rapid con- 

 centration of the water. In 1909 both January and February 

 had a very low rainfall, while April and May were exceptionally 

 snnny; the water-level fell very rapidly in the flrst two months 

 of the year, rose again in Mardi (3-42 in. of rain) and then once 

 more began to diminish. It is possible that such sudden changes 

 in concentration, whether tending towards dilution or the re- 

 verse, affect the Chaetophora adversely. There is further évi- 

 dence for this view in the early disappearance of the genus in 

 1906 and 1909 ; in both years Chaetophora had practically dis- 

 appeared by June. This month in both years was characterised 

 by heavy rainfall, folio wing on relatively scanty rainfall in the 

 two previous months. In 1909 this heavier rainfall continuée! 

 uniformly and Chaetophora recuperafed to aslight extent; but 

 in 1906 the rainy month of June was foïlowed by an alinost 

 rainless July, and this repeated change in concentration of the 

 water may hâve been responsible for the complète absence of 

 the genus during the ensuing months, although it reappeared 

 temporarily with the dilution of the water by the heavy rains 

 of OctoberlOoc». 



Tiible to show occurrence of Chaetophora ht Barton's 

 pond, 1906-1909 (1 >. . 



(1) Regarding the symbols, see footnote 1 on p. 54. 



