- 56 — 



data. The rainfall from the previous October to the May (1) of 



the years in question was as follows : 

 October ( 1905-1906 1906-1907 1907-1908 1908 190 9 

 to May. 15-7 in. 21*25 in. 21-68 in. 12-64 in. 



Table to show the occurrence of Desmids in 

 Bartoris pond, 1906-1900 l2 >. 



Dei 



1906. 



1 1907. 



1908. 



1909. 



1906. 

 J 1907. 

 1 ) 1908. 



1909. 



1906. 

 1907. 

 1908. 

 1909. 



vr. 

 rr." 

 vr. 

 vr. 



vr. 



vr. 



vr. 

 rc* 

 rc* 



— vr. 



rc. 

 rc* 



vr. 

 rr.* 

 rr. 



vr. 



i. 

 r. 

 rc. 



The amountof bright sunshine does not appear to hâve varied 

 appreciably in the différent periods (althouglï rather high from 

 Oct. 1908 to May ly09), nor is there anything in the tempera- 

 ture-data to afford an explanation. The rainfall data however 

 sliow a definite relation to the amount of Desmids subsequently 

 présent in the pond. It is the expérience of everybody, who lias 



(1) The period up to May lias been chosen for thèse data,because the Desmids 

 generally hecomemost abundant early in .lune. The relative amounts of rainfall 

 would however remain much the same, il .lune was included in the calculation 

 or May omitted. 



(2) Regarding the symhols, cf. l'ootnote 1 on p. 54. An ' indicates fréquent 

 végétative propagation. 



