October, 18'Jl.l 257 



afterwards cuttiiii^ the pocket without the citnstjint risk of the point 

 coming tlirough the wall. IIow the disqualilicatiou is overcome T 

 hope to be able to show. 



Having kept ixsemipurpurclla in confinement, I had the advantage 

 on several occasions of watching her method of proceeding. As the 

 plates of the last (7th) abdominal segment gradually opened, and the 

 ventral one was at the same time extended and advanced within the 

 lip of the pocket, it was easy to see, through the transparent mem- 

 brane filling the gap between them, the spurs of one side gliding up 

 and down, the lower spur appearing more erect than in the quiescent 

 state, and standing nearly at right angles to the line of the instrument. 

 As the depth of the pocket increased, the spurs passed out of sight, but 

 would presently flash by again, followed by a brpwn body — the knife 

 and its rolled up sheath, and would disappear within the abdomen. 

 Again would the instrument be shot down into the pocket, where it 

 would remain for some seconds, and then once more flash into sight 

 and disappear as before ; and this would be repeated several times 

 before it was finally withdrawn, slowly and cautiously, as compared 

 with the former rapidity, so as not to scratch nor injure in its passage 

 the thin-shelled and delicate egg. But if, instead of this side view, 

 we take a full view and look through the upper surface of the leaf, 

 we shall see that the saw does not make rapid alternating cuts from 

 side to side, but that it works away for a time at one wall, and is then 

 withdrawn for an instant, and on starting again begins on the other 

 wall, and that it is in a sort of piece by piece fashion that the pocket 

 is cut. At the same time a singular feature is noticed, which is, that 

 the knife plate loses in symmetry, bulging out on the one side, viz., 

 that on which it is working, and becoming slightly concave on the 

 other side. It might be thought that this is merely the effect of the 

 resistance it meets with, just as a stick bends when it is pressed against 

 an object, but this can scarcely be the true explanation, since the same 

 shape is sometimes seen in specimens under the microscope (Fig. 0, 2), 

 where no such resistance can be present, and we are, therefore, 

 forced to the conclusion that the power to alter its form, unlikely 

 as it may at first sight appear, must be inherent in the instrument 

 itself. 



Now, I wish to draw particular attention to one point, not 

 apparently a very large one, in the above observations, namely, the 



13 u 



