MACROTHYLACIA RUBL 135 



observes that on Baildon Moor, near Bradford, the larvae occurred 

 in such profusion in 1874 that one could scarcely walk without 

 treading them under foot. A similar abundance was observed in 

 September and October, 1894. Ash notes that the larvae were 

 also excessively abundant in the latter year on Skipwith Common. 

 In 1856, the larvae were reported as occurring at Bisterne in 

 marvellous numbers, strewing the ground ; indeed they were in such 

 numbers that a person walking across the heath known as the 

 " warren " had to be cautious lest he should tread on them. 

 Home reports them as being in thousands in most years on the 

 high ground around Aberdeen, and Wylie that, in certain seasons, 

 they are extremely abundant on the mosses in Perthshire ; in the 

 autumn of 1897 the larvae were feeding on nearly everything on 

 Methven Moss. McArthur says that in the Isle of Lewis the 

 larvae were in amazing abundance about the middle of September, 

 about three parts grown and almost everywhere. The larvae 

 swarm in Sketty Park, where there is no heath, and were so 

 abundant in September, 1891, that it was difficult to avoid 

 treading on them (Robertson) ; unusually abundant the same 

 year at Clevedon, where, also, there is no heath, the larvae feeding 

 on grass till about three-quarters grown, when they collect in small 

 colonies on bramble until fullfed, and then enter into their winter- 

 quarters ; the larvae will feed well on bramble in confinement, but 

 also eat all the grass in their enclosure (Mason). In spite of the 

 usual autumnal abundance of the larvae, there is rarely any great 

 number found in the spring, and one suspects that large numbers 

 perish during the hybernating period. All lepidopterists complain of the 

 difficulty of rearing the species if kept in confinement throughout the 

 winter, and thus it is usual for British lepidopterists to force them. 

 Cartmel appears to have been the first to adopt the now well- 

 known method of forcing the larvae by bringing them indoors in 

 winter and placing them near a kitchen fire. By this means he 

 succeeded in breeding many imagines in January, 1856, from larvae 

 that he had collected fullfed the preceding October. Robson places 

 each fullfed larva in a paper box at least 2 inches square (chip 

 boxes are too small and result in cripples), in early spring, and 

 gives {E.M.M., xxxiii., p. 199) a long detailed account as to how 

 to rear the species, the larvae being collected from January 10th- 

 March 13th, dependent on season, when they have come up to sun 

 themselves. He places them, at once, as we have just noted, into 

 a separate paper box (these paper boxes being kept on a shelf 

 over the kitchen fire), in which the larvae almost immediately com- 

 mence to spin up and are thus forced so as to emerge early, e.g., 

 larvae February ioth-i3th, 1897, spun cocoons February i4th-2oth, 

 imagines emerged April 4th on. It is remarkable that, in some cases, 

 the forcing appears to hasten the assumption of the pupal period, but 

 usually to lengthen the pupal stage. Thus, larvae, forced from February 

 ioth-i3th, 1897, produced imagines in an average of 58^ days (the 

 longest period 64 days, the shortest 53). Those found March 21st- 

 April nth, 1897, produced imagines in an average of 28 days (the 

 longest period being 35, the shortest 21). In all cases the first to 

 appear were $ s, the males appearing three or four days later. With 

 regard to this point, viz., that the earlier found larvae, though pupating 



