OCCASIONAL ABUNDANCE OF INSECTS. 39 
phenomena I am investigating are the result of causes 
operative during one season and not neutralized till the 
end of the next. Besides this the last series 1887, 1888 
seems to uptset any conclusion drawn from a high tem- 
perature as although the July, August of 1887 were 
considerably above the mean, the same months in 1888 
were distinctly below it. 
Now allow for the sake of argument that the hot 
summer of 1887 originated the abundance, we must then 
admit that the very cold summer of 1888 would have 
checked it, and checked the survivals before they reached 
the adult larval stage in which they were discovered so 
numerously. But the problem to be solved provides that 
the check came in when or after the larve had reached 
maturity. Therefore the temperature factor will not solve 
it. If temperature really had any great effect the series 
1868, 1869 or 1884, 1885 should have been ‘ galil”’ years 
instead of 1869, 1870 and 1887, 1888. 
T admit to some disappointment at this want of result 
as I felt that a cold summer ought to have had a very 
marked effect in retarding and a hot one an equal effect 
in stiinulating survivals, affecting as I assumed it might 
both the summer oviposition and the welfare of the very 
young larve. Thus foiled in my attempt to deduce any 
decent theory out of Temperature I turned to Rainfall. 
Now here I do flatter myself that I can trace some sort of 
principle at work. The lowest figures give the mean of 
the half century. 
And firstly one thing is obvious, the imagines are no 
more effected by rain than they were by temperature. In 
June, July, 1888, 5°05 inches of rain fell, while the mean 
is only 2°21 and the two summers of the biennial 1887-8, 
giving altogether nearly seven inches of rain resulted in 
more larve than 1869, 1870 when the two summers give 
