CLASSIFICATION 353 



resisted the temptation to divide the large and diverse genus Actino- 

 myces until a real basis for this division became evident. The fol- 

 lowing key taken from their recent work has in it ideas from several 

 earlier workers, 0rskov, Jensen, Krassilnikov," Brumpt, and others. 

 This reclassification of the actinomycetes was the last paper sent in 

 for publication by Henrici before his untimely death. 



KEY TO THE GENERA OF THE ORDER ACTINOMYCETALES 

 (Adapted from Waksman and Henrici '") 



A. Mycelium rudimentary or absent. Family MYCOBACTERIACEAE 



Genus Mycobacterium 



B. Branched mycelium produced. 



1. Vegetative mycelium divides by segmentation into bacillary or coccoid ar- 



throspores. Conidia not produced. Family ACTINOMYCETACEAE 



a. Anaerobic or microaerophilic, usually parasites of animals, not acidfast. 



Genus Actinomyces 



b. Aerobic, partially acidfast or not acidfast. Sometimes pathogenic to 

 animals. Genus Nocardia 



2. Vegetative mycelium normally not divided into arthrospores. Conidia pro- 

 duced on proper media. Rarely if ever pathogenic to animals. 



Family STREPTOMYCETACEAE 



a. Conidia found in chains from aerial hyphae. 



Genus Streptomyces 



b. Conidia formed terminally singly or in small clusters on conidiophores, 

 not in chains. Genus Micromonospora 



The above system is very similar to an earlier classification by 

 Puntoni. It differs in being more complete and in that International 

 Rules of Nomenclature are followed. Concise but lucid Latin de- 

 scriptions of Puntoni 's three genera in a paper,"^ which in effect sum- 

 marizes his work, shows how closely he anticipated the system above. 

 In our opinion, however, none of the generic names used by Puntoni 

 is valid as used. 



Concerning the origin and evolution of the actinomycetes, three 

 possibilities have been considered: (1) that they are a higher de- 

 velopment of the bacteria; (2) that they are degraded molds; and 

 (3) that they represent a common ancestral type from which both 

 bacteria and molds have developed. We hardly know enough about 

 them yet to warrant such speculation, but all three theories have 

 had their supporters. Lieske " leans toward the third viewpoint be- 

 cause of the fact that the actinomycetes are very labile and prone 

 to vary in their morphology, at times toward the bacteria, at other 

 tjmes toward the molds. The following scheme taken /rpnj his book 



