4:74 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



chitinous ridge, the latter, except at the extremities, being thrown 

 into a series of deep curving folds. This dorsal groove curves down 

 to the mantle edge rather abruptly at the hinder end, but more 

 gradually at the anterior extremity. Round the edge of the 

 mantle there are 24 dark brown spots, many of which are extended 

 toward the dorsal region as irregular marks, and between these 

 markings there are sometimes smaller spots of the same color. 



On each side there are nine spiracles. The first lies just above the 

 second dark spot of the mantle edge, the second is above the fourth 

 spot, and the remainder correspond in position to the mantle spots 

 beneath them. The third is placed very far up the side, the fourth 

 is a little lower, and the remainder are still lower and in a nearly 

 straight line. 



The whole structure of the exposed portion of this remarkable 

 larva is profoundly modified, presumably as a protection against 

 the attacks of the ants by which in life it is surrounded. 



The entire skin is covered with thick chitinous plates, which are 

 irregularly radiate and have sloping edges. The projections of these 

 plates interlock with the lateral cavities of those adjacent, and the 

 appearance luider a high power is not unlike that of a complicated 

 armor sheet of cogwheels. Such an arrangement doubtless pro- 

 vides a very efficient protective covering with a maximum of flex- 

 ibility. The plates vary considerably in size and in the extent to 

 which the edges are sloped. 



Microscopic examination shows that the brown markings on the 

 larva are caused by the presence of small brush-like chitinous 

 tufts, one of which arises from the socket in the center of each chitinous 

 plate. In the unpigmented parts of the skin, from which these 

 tufts are absent, the plates still have the sockets, so that at one period 

 of its evolution the larva probably had the tufts, or at least some 

 scale-like grow^th, on every plate. 



Here and there in the armor, especially (probably exclusively) 

 on the pigmented areas, there are small round openings, the edges 

 of the adjacent chitinous plates being neatly hollowed, so that each 

 forms its representative part of the circle. Doctor Eltringliam sug- 

 gests that possibly these apertures are the external openings of 

 glands, though he had no proof of this. If, as he believes, they 

 are confined to the pigmented areas, this would probably account 

 for the correlated presence of the brush-like tufts, which may either 

 protect the openings, or, as Professor Poulton has suggested, hold 

 some attractive secretion prized by the ants. 



The extreme outer edge of the mantle is armed with a regular 

 fringe of flat, chitinous projections, their bases furnished with inter- 

 locking processes. On the upper side of each projection (lieie is a 



