122 • f^^-''^' 



course, for a wbile the sexes were coming out simultaneousl^^ Tlie 

 individuals of botli sexes vary surprisingly in size, the males from 9 

 to 13"5 mm , and the females from 8 to 13 mm., and all intermediate 

 sizes are well represented. Some specimens are rather more grey, 

 with the black spots more numerous and larger, wbile others are 

 somewhat whiter, or more ochreous, with the black spots fewer and 

 smaller, but the slight variation to be seen in colour and marking is 

 not sexual. The ovipositor of the female is of considerable length, 

 doubtless to enable it to hide its eggs away safely in the middle of the 

 seed-head, and in dead specimens is generally more or less extruded. 



The full-fed larva hibernates, licad upwards, in its mine inside the seed -hearing; 

 stem, and in the spring it pupates therein, as suggested by Lord Walsingham. 

 Being anxious to learn more of its habits, I lately opened up numbers of the stems 

 (a most laborious task) until I had collected many larval mines, fully a score of 

 which contained pupa-shells enclosed in cocoons. The mine inside the stem some- 

 times extends downwards a little below the level of the lowest seed-vepsels, but it 

 was never found to reach more than 13 mm. below this point, and all the cocoons 

 were above it, some being only just above, while probably all were within 20 mm. 

 of it. Since the stem is too narrow to contain the pupa, the larva widens that part 

 of its mine, which has been chosen for the site of the cocoon, by gnawing away, (o 

 a greater or loss extent according to the necessities of the case, the walls of the 

 stem and of the adjacent seed-vessels: then, in the wider chamber thus formed, it 

 spins a tougli cocoon, the walls of which, esce]it where the natural stem walls arc 

 left intact and utilized, are composed of gnawed up vegetable fibre mixed with 

 cement, and nearly resemble, both internally and externally, the natural stem-walls. 

 The whole of the interior of the cocoon is thinly lined with white silk, which looks 

 glossy and hard as if it had been besmeared with gum ; its surface, however, is not 

 very smooth and even. In its nature and composition, tlie cocoon, which varies 

 considerably in size, but would average about G"2 mm. long, by 1'4 mn^ broad at its 

 widest pai't, reminds one strongly of that of a Centra. The upper end of it always 

 occupies the interior of a hollowed-out seed-capsule, through the top of which, 

 apparently weakened by the larva, the imago escapes, leaving tlie whole of the pupa- 

 shell behind it, inside the cocoon. Tlie remainder of the cocoon lies in the enlarged 

 larval mine, except when, as occasionally happens, the lower end, which always has 

 a pad of white silk threads spun across it, occupies another seed-vessel. The pupa, 

 as shown by the position of the empty shell, stands upright with its anal extremity, 

 surrounded by the east larval skin, resting on this silken )iad. In one solitary 

 instance, among fully a score of cocoons that were examined, the whole cocoon lay 

 outside the stem, in a chamber parallel to, and formed against, the wall of the stem, 

 by the larva having gnawed away part of the substance of the wall, and parts of 

 three adjacent seed-vessels. In another single instance, an abnormally thick stem 

 was wide enough in itself to contain the cocoon, the upper end of which, however, 

 occupied, as usual, a seed-capsule. 



The following brief description was made from the empty pupa-shells. These 

 will be sent to my esteemed friend. Dr. T. A. Chapman, in the hope that they may 

 afford something of interest to the specialist. 



