210 Rev. G. A. Crawshay on the 
I had tried various others without success, and had little 
hope of ascertaining the number of moults the larva 
undergoes. I had, moreover, exhausted all my ova with 
the exception of one late batch, for I did not then contem- 
plate a second brood of beetles emerging and ovipositing 
as late as September. 
Early in July 1906 with eighteen newly-hatched larve 
I made my last experiment, by the flat glass method, and: 
of these I only succeeded in bringing six over the 8rd 
moult, four over the 7th, one over the 8th, and none to 
perfection (ante, p. 201). 
Hitherto I have dealt with the feeding up of the larva 
in bark. When full fed it is advisable to transfer them to 
wood. In November 1905 following the June of that 
year when I first discovered this beetle I was confronted 
with a number of larve apparently full fed in a tree on 
which my females had laid in June. Some of these 
appeared to me, in this incredibly short time for a 
Longicorn larva, to have actually established themselves 
in their pupa-cells and seemingly would hibernate thus. 
No. 3.—Accordingly I devised a method of bringing 
these through to perfection quickly in wood. In this way 
I obtained imagines almost at once, and was able to watch, 
for the first time, the transformation to the pupal and 
perfect states under glass. 
I used the ?-in. corked glass tube for the first time. 
Let into these were solid cylinders of fresh larchwood 
turned on a lathe exactly to fit the tube, leaving no space 
for mildew (Plate XIX, «a, b,¢,d). In the top of each piece 
of wood with a }-in. gouge I made grooves in the outer 
surface 14 in. long, extending downwards, large enough 
comfortably to take the larva and leave it space for packing 
away its rejected wood-fibre in forming the pupa-cell here. 
Placed in these grooves the larvae smooths down the 
interior walls, shaping them to its liking, and forming a 
pupa-cavity, as in the normal tree, except that the glass is 
used for one side. Sometimes it continues the burrow 
downwards, thus affording an excellent opportunity of 
observing the mode of excavation and the ramming 
process that takes place in the natural tree. 
Tn each tube four to five larvee were fitted up in adjacent 
grooves, and every movement could be observed through 
the glass. 
