139 



throughout the day visiting the flowers in gardens, fighting in pairs 

 in the streets, often being carried away on the wind, but chiefly 

 congregating in favoured rough corners and lanes, and the banks of 

 the railway, where the early morning sun seems especially to have 

 drawn them out into full flight before 9 a.m. 



Such a numerical increase in these two emergences suggested the 

 probability of a corresponding increase in the progeny of the 

 third brood, but J. was not at all prepared for what I actually 

 saw on September 23rd. A grassy lane between pasture-land 

 which is dry and open, besides many species of grasses, is rich 

 in Lotus, Galium, Cardamine, etc., etc., and especially fertile in 

 species of Ruinex, common sorrel {Rumex acetosa) growing in 

 quantity. Here, during this month (September), P. phlcsas has 

 revelled. The first root inspected showed about a dozen of the ova 

 placed anywhere on the leaf, both upper and under sides, and even 

 along the stalks where exposed. Succeeding clumps were equally 

 patronized, and in some' cases individual leaves of the plant were 

 literally besieged with eggs, as when a withered leaf on a long stalk 

 projecting on to the footpath, was found to have twenty-one eggs 

 and four young larvae attached to it. Altogether, in an hour I 

 selected more than 100 of the ova, and left at least twice as many 

 for future observation of the larva. I should at once mention that 

 this abundance was, in a manner of speaking, quite local. One side 

 of the grassy lane faces the south east and catches the early morn- 

 ing sun ; there is no ditch, and the sorrel and dock grow quite 

 commonly under a hedge in a dry gravelly soil. Although dock was 

 equally plentiful with sorrel, the latter plant was always preferred, 

 indeed, on only one clump of dock did I find ova and larvae, and 

 then on a young and stiff-leaved plant. The luxurious growths of 

 common dock were entirely neglected, as also were both species of 

 Rumex when growing amongst the turf, or in a damp ditch running 

 the length of the opposite side of the lane. This latter fact points 

 decisively to the cause of variation in numbers, and it is quite clear 

 that this summer the same conditions have acted with exacily 

 opposite effects on species like Pieris napi and P. phlceas — with the 

 former entirely delaying a second brood except in the dampest 

 districts ; with the latter, P. phlceas, enormously increasing, and 

 palpably accelerating the emergence of two successive broods, A 

 wet autumn and winter will doubtless thin down the [)resent 

 abundance to a normal spring emergence in 1894 ; while a dry spell 

 in January, February, and March, will do much to preserve an 

 exceptional quantity of the larvae, and produce a large brood during 

 next May and June. It may even happen that, provided October of 

 this year be reasonably warm, a partial fourth brood will put in an 

 appearance at the end of that month ; and thus in 1893 four broods 

 will have appeared in the period ordinarily taken to produce three. 

 I took a good female last year on October 31st, after several nights 

 of frost. 



