2i8 PROTOBASIDIOMYCETES [ch. 



species has not been investigated, but there is an indication of a transition 

 between Endophyllum and the eu- forms. It is postulated that the cells of 

 the promycelium of an Endqphyllum-Wke: ancestor might have produced 

 infection directly instead of by means of basidiospores. Nuclear fusion and 

 meiosis would be postponed till the formation of teleutospores, which would 

 arise as the typical spores of this new vegetative sporophyte. Later the 

 teleutospore would become specialized as a resting spore and uredospores 

 would take on the function of rapid propagation of the plant. 



But Endophyllum has a definite cluster-cup, and well-marked peridium, 

 and is therefore not likely to be primitive, though it may point to an 

 hypothetical primitive ancestor ; a species must rather be sought with a 

 caeoma from the aecidiospores of which promycelia are produced. Such 

 a form has been recognized in Kunkelia nitens. 



There remain to be considered the various forms of teleutospore. 

 Presumably the unicellular type is more primitive than the multicellular. 

 Simple teleutospores occur in Melampsora (where also the aecidiospores 

 are developed in a caeoma), in Uromyces and in the Coleosporiaceae. In 

 the last-named family meiosis and septation take place inside the teleuto- 

 spore wall, but the elaboration of the other spore-forms forbids this group 

 being regarded as primitive, though the internal basidium may be. 



It may be hazarded that the ancestral rust bore spermogonia or groups 

 of male organs and g-roups of female organs of the caeoma type, that the 

 individual male organ was an antheridium or spermatium set free from its 

 parent hypha, that the female organ consisted of a fertile cell or oogonium 

 and a sterile cell which was perhaps elongated to form a trichogyne reaching 

 up to the surface of the host, and that the product of fertilization was 

 a series of aecidiospores. It may further be suggested that either the 

 aecidiospore germinated by giving rise at- once to a promycelium or that 

 an alternation of vegetative generations occurred and that the sporophyte 

 bore simple teleutospores or tetrasporangia inside which septation took 

 place. 



The members of the Uredinales may be arranged in four families : 



Germination by a promycelium (except in Chrysopsora) 



Teleutospores stalked Pucciniaceae. 



Teleutospores sessile 

 arranged in series but separating later Cronartiaceae, 



one or many celled, loose in tissue of host or 

 united in a flat layer under the epidermis Mebampsoraceae. 



Germination without a promycehum, formation of basi- 

 dium internal; teleutospores sessile or with a lateral 

 pedicel Coleosporiaceae. 



