THE MONTHLY BULLETIN. 303 



tightly closed at first, they open up when about half grown so as tn permit entrance 

 of nrodntorv insects, r.nrio numbers of flip spring migrants become the prey of 

 spiders, which are often abundant about the elm tree. 



Judging from the number of trees planted and flourishing today the rork elm has 

 been a favorite shade tree along roads, nhout ranch buildings and in towns in 

 California. The aphis has thus been furnished with propitious conditions for its 

 full development. There is little doubt that hoth spring and fall migrants can 

 migrate comparatively long distances. The writer has found both forms well over 

 a mile from their respective localities of origin and feels reasonably sure that they 

 can travel much greater distances. Both forms are somewhat elongate, hare, dark 

 green or hrown. the disc of the abdomen shining and the body tipped with a small 

 tuft of white "wool." 



The Root-Feeding Form. 



Relative ahundaitrr thrnvnhnwt the year. — During the winter months from 

 December to March the aphis is scarce, occurring mostly on fibers. On trees 

 which have been heavily infested the previous summer it is frequently hard to locate 

 any aphis in winter. In April and May the numbers increase regularly. Occa- 

 sionally, however, on lighter soils quite heavy infestations are encountered early in 

 April. After May the increase in numbers is more rapid until September when 

 the maximum numbers occur. Tn this month there is the greatest production of 

 winged aphids and thereafter the numbers dwindle rapidly, although it happens at 

 times that the production of migrants and maximum infestation is postponed as 

 late as the end of October. As a rule relatively more migrants are produced in 

 heavy than in light soils, and. therefore, in the latter there is a less noticeable 

 diminishing in numbers, following the development of the migrants. 



liinlnpir observations. — To secure laboratory data on the reproduction, develop- 

 ment and habits of the root aphids so that conditions might approach those of 

 nature the insects were bred on root sections kept in moist sand in a cellar. 

 Almost invariably the aphids if disturbed moved off and refused to settle again on 

 the desired piece of root and. therefore, unless they were let alone records could not 

 be secured. This resulted in manv incomplete records, as the small roots dried up 

 and the insects departed before full development or before the deposition of the 

 full ouota of voting. Individual reproduction varied from 90 young in 33 davs to 

 IS young in 28 davs. The average number deposited by an individual was 40 and 

 the dailv average I. ft. yet on several occasions 7 and in one 8 young were extruded 

 within 24 hours. Toward the end of the deposition period days were frequent on 

 which no young were deposited, and this explains the small average daily repro- 

 duction. 



The newly-hatched root nobis is elongate in shape, yellowish pink in color, and 

 hare. After a few hours' feeding rows of little white wax tubes appear over the 

 hodv. These grow into hollow filaments and attain their maximum length not 

 until several days have el-insed. The wax filaments or threads in the first instar 

 mav become as much as eisrht times the length of the aphis. Just preceding each 

 molt these threads break off at the base and after the skin is molted the body is 

 ouite bare and the threads crow out anew. Tn the later instars the threads are 

 never over three times the length of the insect's body. Four molts occur before the 

 aphis is mature and after each molt the color of the newly-molted individual is 

 darker than at the one previous so that the adult insect is at first reddish brown. 

 It is elongate pyriform in shape and about two mm. in length and one mm. maximum 

 width. The body color soon becomes pink, and later, preceding the insect's death, 

 turns dark red. Not until four or five days after the molt do the wax threads 

 attain their full length, which is barely twicp the length of the aphis. The threads 

 have a tendency to curl and split at the end. and when not forecasting a molt (heir 

 breaking off at the base is a sign that the aphis is getting inadequate nourishment 

 or is moribund. 



During the winter months the roof aphis is in a state of virtual hibernation. 

 Specimens in nil stages of growth collected in December showed no activity until 

 February, when the larva? molted and the adults began the denosition of young. 

 The insects pass the winter in any stage, but it is hardly a true hibernation, rather 

 an unusual prolongation of stages and cessation of reproduction. During the 

 snring months the aphids developed in an average of four weeks varying from 

 20 to 3." davs. In June this period was hardly shortened, but in July. August 

 and September the average dropped to 18 days with a minimum of 13. In 



