CHARACTERS OF THE ACTINOMYCES. 285 



of which filaments radiate for a short distance (Fig. 74). 

 The filaments usually stain uniformly in the younger 

 colonies, but some, especially in the older colonies, may be 

 segmented so as to give the appearance of a chain of bacilli 

 or of cocci, though the sheath enclosing them may generally 

 be distinguished. Rod-shaped and spherical forms may 

 also be seen lying free. 



2. Coccus-like Bodies. The formation of these from 

 filaments has already been described, but it is doubtful if 

 all are of the same nature. Like other species of strepto- 

 thrix the actinomyces when growing on a culture medium 

 shows on its surface filaments growing upwards in the air, 

 the protoplasm of which becomes segmented into rounded 

 spores or gonidia. In natural conditions outside the body 

 these gonidia become free and act as new centres by 

 growing out into filaments. They have higher powers of 

 resistance than the filaments though less than the spores 

 of most of the lower bacteria. It is probable that some of 

 the spherical bodies formed within filaments when growing 

 in the tissues have the same significance, i.e. are gonidia, 

 whilst others may be merely the result of degenerative 

 change. Some observers have described young colonies 

 largely composed of spherical forms as if these multiplied 

 by division, but this latter point is still doubtful. Both the 

 filaments and the spherical bodies are readily stained by 

 Gram's method. 



3. Clubs. These are elongated pear - shaped bodies 

 which are seen at the periphery of the colony, and are 

 formed by a sort of hyaline swelling of the sheath around 

 the free extremity of a filament (Fig. 75). They are 

 usually homogeneous and structureless in appearance. In 

 the human subject the clubs are often comparatively fragile 

 structures which are easily broken down, and may some- 

 times be dissolved in water. Sometimes they are well 

 seen when examined in the fresh condition, but in hardened 

 specimens are no longer distinguishable. In specimens 

 stained by Gram's method they are not coloured by the 

 violet, but take readily a contrast stain, such as picric acid, 



