THE TERRAPIN SCALE. 27 



Length of the First Instar. 



Tlie earliest molts were upon the sixteenth day and were observed 

 during the very favorable weather of June and July, 1913. Eighteen 

 days is the average length of the first instar at Mont Alto during fav- 

 orable years, as shown in Table XVIII, columns 7 and 8. This time 

 may be nearly doubled by unfavorable weather. Honeydew is ex- 

 creted during this instar, but in very small quantities, and is of no 

 economic importance. 



Dispersal of First-Instar T^arv^ by Leaves. 



It is probable that this species is dispersed to some extent by the 

 transportation of larvae upon wind-borne leaves during storms. An 

 experiment performed July 22, 1913, showed that first-instar larvae 

 can loosen from slowly drying leaves and that they can move about 

 and reattach to living foliage, so that if infested leaves should lodge 

 in adjacent trees the latter would undoubtedly become infested. 



Sexual Dimorphism in the First Instar. 



There are no noticeable indications of sex during this instar, 

 except in the anal ring. It is possible in some cases to distinguish 

 the females from the males after the fifteenth day by their increased 

 width. At this time the length of the females is usually less than 

 twice their width, while the length of the males is usually greater 

 than twice their width. 



Nearly all specimens are distorted by crowding, or by contact with 

 the veins of the host (fig. 5), so that this variation m the ratio of 

 length to width can not be depended upon for distinguishing the sexes. 

 By dissection, however, they can be distinguished. The anal ring 

 of the male consists of only six setae, while the anal ring of the female 

 consists of eight. 



The First Molt. 



There is no change of position at the first molt. The skin splits 

 along the back and is worked downward and backward underneath 

 the body. The last portion to loosen is that about the anal plates. 

 The major apical setae disappear at this molt; hence the absence of 

 these is positive evidence that the first molt has passed. 



The larvae stop growing one day before molting and become more 

 opaque. The time required to make this molt is from 5 to 30 min- 

 utes, depending upon the weather conditions and the vigor of the 

 larvae. The molt is usually made in the early morning. 



Observations made upon 5,000 larvae approximately one-haK of 

 which emerged from June 24 to August 9, 1912, and the others from 

 June 24 to July 1, 1913, show that this molt may take place as early 

 as the sixteenth day and as late as the twenty-sixth day. The aver- 



