310 



iv, p. 173]. Economic injury by this beetle was first recorded in 1894 

 by Davis, who found it feeding on cultivated cherry in Michigan. 

 The 1915 outbreak was probably the most injurious that has ever 

 occurred and advantage was taken of it to secure as complete data as 

 possible in regard to the natural food-plants of the beetle, its life-history 

 and the means for its control. The increased numbers of G. cavicollis 

 may be attributed to an increase in abundance of its natural food-plant 

 Prunus pennsylvanica {pin-, fir-, or bird- cherry). This tree springs up 

 rapidly along road-sides and in partially cleared or fire-swept forest 

 land which has been left uncultivated. Such lands cover wide areas 

 in western Pennsylvania and furnish ideal breeding conditions for the 

 beetle. The adult feeds almost exclusively on the underside of the 

 leaves of the plants attacked. It also feeds on the fruit of the cherry. 

 Occasionally the beetles have been observed feeding on the upper 

 surface of peach leaves and once on the upper surface of cherry leaves. 

 The preference for sickly or injured trees was marked. Such trees are 

 invariably loaded with beetles, while the surrounding trees may be 

 comparatively free from attack. The foliage on an mihealthy branch 

 was attacked before the rest of the tree. The preference for the foliage 

 on the lower limbs to that on the upper was still more marked, for the 

 former may be completely defoliated while the latter are comparatively 

 uninjured. The period of economic injury extended over 14 or 18 

 days after its first appearance in June, the greater part of the feeding 

 being probably done during the first three or four days. There was 

 no injury noticeable from the later brood. Severe injury was almost 

 entirely confined to the Early Richmond cherry, especially to young 

 trees. All the stages of G. cavicollis are described. After an average 

 feeding period of about 12 days, the larvae entered the ground, remain- 

 ing there for an average period of 22 days, of which the pupal stage 

 occupied about 11 days. The emergence of the adults took place from 

 a day to several days after transformation. The total developmental 

 period lasted from 45 to 50 days. In captivity the sexes appeared in 

 almost equal numbers. The earliest adult appeared on 23rd August, 

 though beetles were observed somewhat earlier in the open. In the 

 leaf-mould at the base of wild cherry trees, in which G. cavicollis were 

 transforming in large numbers, small Carabid beetles were found, which 

 have been determined as a large form of Lebia ornata, Say. In con- 

 finement these Carabids readily attacked the pupae and freshly emerged 

 adults of G. cavicollis. In the course of a number of control experi- 

 ments, arsenate of lead, not less than 5 lb. to 50 U.S. gals, of water, 

 was effective in protecting trees from injury. An addition of 1| U.S. 

 gals, of molasses produced a more effective mixture that killed 

 practically all the beetles on the trees to which it was applied, but this 

 addition of sweetening matter had the serious disadvantage of rendering 

 the spray readily removable by rain. Forty per cent, nicotine sulphate 

 with water at the rate of 1 to 600, and with or without soap, was 

 effective as a contact-spray. Weaker solutions of nicotine sulphate 

 and soap-carbolic acid solutions, though apparently effective at the 

 time of application, did not have a permanent effect. More extensive 

 spraying experiments are required. As the larvae feed on a wild plant 

 where control measures cannot be applied, the clearing up of cut-over 

 woodland and its destruction would greatly diminish the damage 

 done. The underside of the leaves must be covered by the spray, 



