410 Transactions. 



in conjunction with vii, but is absent in ttie other segments, vii is 

 beneath iv in the abdominal segments. (Fig. 1.) 



Larval Habits. 



During the first two stadiums the larva burrows close against the outer 

 cuticle of the leaf, but after the second moult descends at once into the 

 parenchyma, gradually eating all that portion of the leaf between the two 

 outer cuticles. It is easy to see where the moults have taken place, not 

 only from the appearance of the frass, as shown in a former paragraph, 

 but also from the fact that at such places the mine is slightly enlarged and 

 the next portion of the mine is directed at a slight angle from the old one, 

 as if the larva had tried to avoid its cast skin, the head-piece of which can 

 be readily distinguished in the clear portion of the mine. After the second 

 moidt the mine is at first narrower than before, and is difl&cult to follow 

 on account of its descending into the denser portion of the. leaf (see fig. 3). 

 The larva appears to undergo only two moults within the mine. 



The first two stadiums are of equal duration, being about ten days, 

 though this period may be shortened or lengthened according to meteoro- 

 logical conditions. The 3rd stadium is longer, lasting, on an average, a 

 fortnight. When full-grown the larva leaves the leaf and descends to the 

 ground to pupate. 



The Cocoon. 



The cocoons are exceedingly pretty little structures, and their con- 

 struction is most fascinating to watch under low powers of the microscope. 

 The usual length is 7 mm., and the width varies according to the situation 

 of the cocoon ; when constructed on a flat surface it may be as much as 

 5 mm. The shape is oval and flat. The construction is exceedingly deli- 

 cate, the cocoon proper having the appearance of a fine white shiny skin. 

 The term " cocoon proper " is used because when completed the whole 

 exposed surface is thickly covered with numbers of minute white floccy 

 globules. These look like small bubbles and when seen through the 

 microscope each one has the appearance of being a collection of minute 

 transparent cells, the whole forming a somewhat lengthened sphere. These 

 globules are exceedingly delicate, and many appear to be lightly attached 

 to the cocoon, though numbers can be removed by blowing upon them. 

 The object of this elaborate superstructure is not known. 



The globules are ejected from the anal aperture of the larva during 

 the construction of the cocoon proper. The larva takes no notice of the 

 newly ejected globule till it comes across it while weaving the silken 

 canopy ; on finding the globule it roughly tears an opening in the structure 

 it has been taking so much trouble in making, and forces it out through 

 the opening so made, not taking the trouble to mend this ugly rent with 

 any degree of care. Between seventy and eighty globules are ejected in 

 this way from the cocoon before the work is completed, and these give to 

 it a most curious and beautiful appearance. Watching two larvae con- 

 structing their gocoons at the same time, I was astonished to find that the 

 globules were ejected by each at regular intervals of about twenty minutes. 

 Having observed the time of one ejection, I was able to observe subsequent 

 ejections at the minute of the operation throughout the two days during 

 which the cocoons were being constructed. The cocoons are to be found 

 in numbers upon the ground in the dead leaves under the food plant : 

 the smallest and most curved and twisted leaves are the most favoured ; 

 even crannies in the lumps of dirt are taken advantage of, besides blades 

 of grass, and sometimes crannies in the bark of the tree itself. 



