NATURE OF POLYHEDRAL BODIES FOUND IN INSECTS. 371 



denser center differentiated from a somewhat lighter periphery. 

 Sometimes within the bodies concentric layers like those of an 

 onion are observable. Often two polyhedra are seen adhering 

 to one another as if in the act of dividing, but an actual division 

 in a hanging-drop has never been observed. When pressure is 

 applied to the cover glass, the polyhedra crack very readily into 

 a number of pieces, and often without the application of pressure 

 the same fragmentation may be observed to occur somewhat 

 more slowly. In the latter case a notch appears at one side of 

 the polyhedron which gradually lengthens into a line progressing 

 slowly toward the other side, much like the cracking of ice. 

 Usually before the line has completely separated the two halves 

 other lines appear, and soon the entire polyhedron is divided 

 into a number of pieces, which may separate or may stick together 

 in a rosette-like fashion. At no time was anything observed to 

 come out of the polyhedra when they cracked in this manner. 

 If the cover glass is moved while applying a little pressure, one 

 half of the polyhedron may sometimes be folded upon the other 

 half without the cracks appearing, showing that it is composed 

 of a tough substance and is not at all brittle like inorganic crystals. 



The only objects in a fresh preparation with which one could 

 possibly confuse the polyhedra are the fat globules and urate 

 crystals, but with a little practice these may be readily dis- 

 tinguished. Fat globules are perfectly spherical and are there- 

 fore unlike the polyhedral shape of the bodies in question; but 

 when in doubt, Sudan III was used, for in this stain the fat 

 globules become red, while the polyhedra remain colorless. The 

 urate crystals are often more acutely angular or are of an entirely 

 different shape from the polyhedra and are frequently traversed 

 by radiating lines. 



Besides polyhedral bodies, fat globules, and urates, a smear 

 from a newly "wilted" caterpillar contains cellular debris, hairs 

 and pigment granules. The pigment granules must not be con- 

 fused with bacteria, for many of them superficially resemble 

 these organisms very closely. When a preparation is dried, 

 mounted, and examined under oil, the pigment granules of the 

 gipsy moth may easily be confused with small micrococci, 

 owing to the fact that they are usually arranged in pairs. As a 



