THE INSECT VIRUSES 371 



polyhedrosis. According to Steinliaus (1949), Cornalia (1856) and Maestri 

 (1856) were among the first to observe the polyhedra and associate their 

 presence with the disease. Following this a long controversy ensued as to the 

 nature of the polyhedral bodies, various suggestions being put forward, the 

 most popular being that the polyhedra were some kind of organism. In 1907, 

 von Prowazek showed that material from diseased silkworms was still 

 infectious after removal of the polyhedra by passage of several layers of 

 filter paper. This was the first step towards the idea that silkworm jaundice 

 was due to a filterable virus; further proof was forthcoming through the work 

 of Acqua (1918-1919) and Paillot (1924, 1926a,b, 1930). Using dark-ground 

 illumination, Paillot saw numbers of minute granules which he believed to be 

 the agent causing silkworm jaundice. Komarek and Breindl (1924) also 

 showed that minute bodies inside the polyhedra could be seen with the 

 optical microscope. This was confirmed by Bergold (1947), who dissolved the 

 polyhedra with weak alkali and observed the liberated virus rods on the 

 electron microscope. It is now known that the polyhedra are protein crystals 

 with the virus particles as an occluded constituent. Without the virus 

 particles the polyhedra are noninfectious. 



The infectivity of the virus rods from the nuclear polyhedra was shown by 

 Bergold (1947) and Bergold and Friedrich-Freksa (1947). They established 

 three facts: (1) The infectivity of the isolated particles is 1 X 10 -11 to 

 4 X 10~ n gm. per larva; this is several orders higher than that of a polyhedral 

 solution before concentration. (2) The sedimentation constant of a suspension 

 of virus particles calculated by infectivity tests of different sedimentation 

 levels is in good agreement with that obtained by the usual optical method. 

 (3) Serological investigations show that the isolated virus particles are 

 serologically unrelated to the host serum and only slightly related to the low 

 molecular weight polyhedral protein. 



The nuclear polyhedra do not stain with Giemsa solution or iron hematoxy- 

 lin without some previous treatment with acids or alkalis; this was first shown 

 by Escherich and Miyajima (1911). The staining reaction is an important one 

 for differentiating between the nuclear and cytoplasmic polyhedra, since, as 

 we shall see subsequently, the latter stain readily with Giemsa solution with- 

 out any treatment. 



Aqueous or alcoholic solutions of bromophenol blue, with or without 

 mercuric chloride, do not stain nuclear polyhedra in spite of their highly 

 proteinaceous character. However, after 10 minutes of pretreatment in 

 normal hydrochloric acid at 60°C, the polyhedra stain intensely with 

 bromophenol blue, with or without mercuric chloride. After the same pre- 

 treatment the polyhedra stain pink with dilute Giemsa solution. 



Although it is possible to make out the bundles of virus rods inside the 

 polyhedra in unstained preparations, by means of the optical microscope, 



