August, '22] smith: Japanese beetle, larval food habits 307 



forms a cell in the soil slightly larger than its body and feeds on the 

 fine rootlets at the top or bottom of the cell. The grubs usually fol- 

 low the course of the rootlets until these are consumed before attacking 

 others. It is this habit of feeding which has prevented the injury 

 to grass from being extremely serious, since it is only in areas of heavy 

 infestation that many plants are found which have all of their roots 

 destroyed. It also follows that in areas which suffer from drought 

 the injury has been the most noticeable and severe. 



The general movements of the larvae in the soil are vertical, where- 

 as the larvae of Cyclocephala immaculata and Anomala sp., which 

 are abundant in this region, usually feed and move in a direction paral- 

 lel with the surface of the ground. During the seasons when the 

 Popillia japonica larvae are feeding they occur in the soil at depths 

 varying between y^ inch and 3 inches. For a short time before the 

 grubs descend on the approach of cold weather in the autumn and again 

 during the period immediately preceding pupation, the depth at which 

 the various individuals are found is more uniform. 



Thigmotropism 



The larvae are positively thigmotropic to living roots and if these 

 are not available, to stones, sticks, or to the bottom or sides of the 

 breeding cage. The larvae have been found abundantly beneath 

 stones in the field and for a distance of two or three feet from these 

 stones no larvae could be found, although at a distance of five or six 

 feet from the stones there would be twenty to twenty-five larvae to 

 the square yard. In a young peach orchard which was cultivated the 

 previous season and allowed to remain fallow over winter, there were 

 nimierous chickweed plants growing in the Spring of 1921. In the 

 spaces between these plants the ground was bare. Fifty plots, each 

 three square feet in area, were examined where no vegetation occurred 

 and no larvae were found. On the removal of fifty chickweed plants 

 an average of 7 larvae were found at the roots of each plant 



Examination of the Contents of the Fore Intestine 



Dissections were made of a large series of larvae collected in grass 

 sod and the contents of the fore part of the alimentary canal were 

 carefully removed. It was found that the material eaten by the in- 

 sects was composed of small soil particles, fresh plant tissue, and small 

 pieces of plants which were partially decomposed. In order to deter- 

 mine the approximate amount of plant tissue as compared with the 

 mineral matter eaten by the larvae, four samples of about five grams 



