118 CYTOPLASMIC INCLUSIONS 



plasmic granules, which he traced back to macronuclear fragments dur- 

 ing conjugation, interpreting these as mitochondria. Since the staining 

 methods were not very specific and these granules do not stain with 

 Janus green, the statement that these are mitochondria must be accepted 

 with caution. Miller (1937) has endeavored to prove that mitochondria, 

 Golgi bodies, and other cytoplasmic inclusions in A. proteus are "bac- 

 teria spores, fungi, or yeasts, together with indigestible material of 

 certain food organisms." This idea is, of course, similar to that of Wallin 

 and has been so thoroughly criticized by Cowdry and others that it need 

 not be considered here in detail. Miller was not able to culture these 

 cytoplasmic "bacteria" and his main argument seems to be based on the 

 observation that mitochondrial stains and Golgi type impregnations will 

 demonstrate granules in the culture medium. This merely shows that 

 the stains used are not always specific under all conditions, a fact which 

 has been pointed out many times. Miller does not present any evidence 

 which can stand up against the observations and experiments of Mast 

 and Doyle (1935a, 1935b), Holter and Kopac (1937), and Holter and 

 Doyle (1938) on the same species. The mitochondria are specializations 

 of the cell itself, probably in all cases from the cytoplasm, and are neither 

 artifacts nor invaders. 



The composition of mitochondria is still incompletely known in any 

 exact sense, in spite of the large amount of work done on these com- 

 ponents. They have long been thought to be composed of both lipoids^ 

 and proteins, because of their staining reactions and solubilities (Faure- 

 Fremiet, 1910; Hirschler, 1924, 1927). Unfortunately no one has 

 yet repeated in Protozoa the work of Bensley (1937), who isolated 

 mitochondria from liver and was able to make both qualitative and 

 quantitative analyses. These analyses confirmed the cytochemical analysis 

 of lipoid and protein, but instead of the large amounts of phosphatids 

 and so forth, which were predicted from the cytological reactions, he 

 showed that the lipoid is largely neutral fat. Bensley also found that 

 many reactions (for example the osmic-acid reaction) are much weaker 



^ Lipoid is used here, as in most cytological works, in a very general sense. It includes 

 all materials which are soluble in ether, absolute alcohol, and so forth, and which stain 

 with Nile blue sulfate, the sudans, and other fat soluble dyes. Neutral fat, fatty acid, 

 phosphatids, and the like respond to these tests. Lison (1936) suggests the rather 

 awkward term sudanophil material, in order to emphasize the cytological side and to 

 avoid false implications as to chemical nature. 



