io6 Journal of Agricultural Research voi. iv. no. 2 



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 micro- 

 cocci, owing to the fact that they are usually arrayed in pairs. As a 

 matter of fact, a smear made from a recently wilted caterpillar is almost 

 devoid of bacteria, and in many cases none at all can be found. If 

 bacteria are found, they have escaped into the body cavity through rup- 

 ture of the intestine and bear no direct etiological relation to the disease, 

 as will be shown later. The very minute, double, dancing granules in a 

 fresh smear, apparently neither pigment nor bacteria, will be treated 

 in more detail in later pages. 



In fixed and stained smears a number of things can be demonstrated 

 to advantage within the polyhedra. Fixation was accomplished either by 

 passing the preparation through a flame or by placing it in absolute 

 alcohol for a few minutes. The smears were then stained in Giemsa's 

 solution for 12 hours or were stained for a shorter period with one of 

 the following dyes: Methylene blue, trypan blue, gentian violet, carbol 

 f uchsin, Bismarck brown, or iron hgematoxylin. When iron haematoxylin 

 was used, the preparation was first mordanted in a 4 per cent ferric-alum 

 solution for two or three hours. Gram's method of staining, Moeller's 

 spore stain, and Welch's capsule stain were also tried, but these were 

 of no greater advantage than the simpler ones given above. After 

 staining, the preparations were sometimes quickly passed through the 

 alcohols to xylol before mounting. This not only clears everything, but 

 dissolves away all the fat, thus increasing the transparency of the prepara- 

 tion. The polyhedra are very resistant to stains in general and usually 

 color along the periphery only, unless the stain is appUed for a long time. 

 On so doing one can succeed in staining the entire polyhedron, especially 

 after the use of some mordant like ferric alum before haematoxylin or 

 anilin water before gentian violet. Steaming the preparation with a 

 stain like carbol fuchsin has also given good results. 



When properly stained, one of three conditions is obtained: First, 

 the polyhedral bodies are uniformly stained so that nothing can be 

 detected within them; or, second, a uniformly darker staining central 

 mass can easily be differentiated from an almost unstained outer sub- 

 stance (PI. XIII, fig. 19); or, third, many little refractive, reddish gran- 

 ules are seen within the polyhedra (PI. XIII, fig. 20). I have obtained 

 very good preparations of these three conditions, especially by the use 

 of iron haematoxylin, methylene blue, and gentian violet. An actual 

 differentiation between what might be interpreted as nuclear and cyto- 

 plasmic material within the polyhedra never occurs; therefore, in account- 

 ing for the staining reactions the writer believes that at times the poly- 

 hedra have a central granular or homogeneous substance easily dis- 

 tinguishable from an outer substance which is more resistant to the dyes. 

 This varies a great deal, however, for very often the periphery of the 



