— 87 — 



en. 's the probability of other factors coming inlo play, such as 

 the gréai concentration of thewater, which took place in April 

 L906 and aothing fcoparallel which occurred inl907and 1908. Il, 

 mav also be noticed that the earlier deci'easeof Peridinhim in 

 1 ( .»07 as contrasted with 1908 corresponds to the very small rain- 

 fallofSeptember 1907, thattheslight increasein November 1906 

 corresponds to verj heavy rainfall in Octobcr and November, 

 and that the « very rare » of November 1908 (as compared with 

 the •• rare » of tlie two preceding vears) corresponds to an 

 exceptionally low rainfall in that mon th. The data however 

 are scarcely adéquate to make sure of this point, but we are cer- 

 tainlv inclined to assume that, apart from warm température, 

 relatively dilutewater alsofavours the development of this spec- 

 ies. The représentation in 1909 seems quite anomalous and is 

 possibly to be explainecl by the présence of the ducks. 



Table to show occurrence of Peridinhim in Bartoris 

 pond, 1906-1909 cl >. 



K. — The Diatoms. 



Among the Diatoms found in Barton's \)(m&Eunotia{Himan- 

 tidium) Arcus, Synedra radians and Eunotia lunaris are 

 the onlyones that play a really important part, and consequently 

 the subséquent considération will be chiefly concerned with 

 them. Of the species of Navicula, only N. lanceolata and 

 N. exilis are at ail common, although the other species men- 

 tioned on p. 43 (incl. the Stauroneis) are almost always pré- 

 sent (but in small quantity). Apart from Synedra radians 

 and Eunotia lunaris the epiphytic Diatoms are of very little 

 importance, although species of Gomphonema are rather 

 constantly represented. 



(1) Reganling the symbols, cf. footnote 1 on p. 54. 



