22 
William J. Hodgetts 
in August—September 1920) does not appear to be quite so harmful 
to Cylindrospermum as to Anabcena. No relation to bright sunshine is 
apparent. Spore-formation in this species regularly follows vegetative 
growth, even when this is very slight (September—October 1919). 
All the species of Oscillatoria and Phormidium which have 
occurred in the pond during the period of observation exhibit a very 
similar mode of occurrence, and for that reason they have been 
grouped together and their frequencies considered as a whole in 
Fig. 11. As is obvious from the curve these genera are present in 
quantity only from June to November, although the actual positions 
of the maxima differ markedly in different years. In each of the three 
complete years there were two maxima, which in the case of both 
1918 and 1919 were due to the development of different species, but 
in 1920 one of the species ( 0 . tenuis ) of the first maximum was 
repeated in the second. Thus in 1918 the first maximum (August) 
consisted of 0 . tenuis and 0 . splendida (approximately equally 
developed), while the second maximum (October) consisted almost 
entirely of Ph. uncinatum; in 1919 the first maximum (August) was 
due to 0 . tenuis (while 0 . splendida was rare), and the second 
(October—November) was caused by Ph. tenue and 0 . splendida, the 
former being rather the commoner species. The first maximum of 
1920 (June—July) was due to 0 . tenuis and Ph. molle in equal 
abundance, while the second (October) was the result of the growth 
of 0 . tenuis 1 and 0 . splendida, the former being rather better 
developed than the latter. Two species (0 . tenuis and Ph. molle) were 
concerned in the rather early development shown in 1921. 
A profound relation is shown between the growth of these two 
genera in the pond and the concentration of the water (cf. data in 
Fig. 3), and in almost every case the position of each maximum 
coincides with the attainment of a high concentration. This is 
naturally most marked in 1919 (the year in which these genera were 
best developed in the pond) when the very high concentration of 
August, and the still higher ones of October—November (cf. Fig. 3) 
exactty correspond with the two maxima of the Oscillatoriaceae in 
this year. Of the six chief maxima shown by Oscillatoria and 
Phormidium two (August, October—November 1919) coincided with 
very high concentrations (27 to 30 degrees), two (August 1918, 
June—July 1920) correspond with moderately high concentrations 
1 Fritsch (14) found that O. tenuis in a pond at Telscombe, near Newhaven, 
had its maximum in October—November, thus agreeing with the present 
observations. 
