612 COMPARATIVE MORPHOLOGY OF FUNGI 



This point of view which has been developed in regard to the Ustila- 

 ginaceae may be carried over directly to the Tilletiaceae. Here there 

 can be no doubt that we have in the structures indicated as sporidia, the 

 remains of basidiospores. Their survival lies in the fact that in the 

 Tilletiaceae, in contrast to the Ustilaginaceae, the mycelial growth by 

 sprouting has much receded or is entirely lacking, hence the reasons 

 given above for the suppression of the basidiospores no longer hold. 

 This original position of the Tilletiaceae is suggested by the fact that in 

 them, e.g., Tilletia, Tuburcinia and Entyloma, the mycelium still cuts off 

 true conidia whereas these are absent in the Ustilaginaceae. On the 

 other hand the basidia of the Tilletiaceae have lost the otherwise most 

 characteristic points of basidia, the constancy of the spore number and 

 the occurrence of the nuclear divisions in them. The nuclear divisions 

 are, as in individual Ustilaginaceae, shifted forward into the smut spores 

 so that in Neovossia the spore number may multiply. The next question 

 is whether and how far these phenomena can be explained by biological 

 influences. Neovossia suggests many Discomycetes in which the spore 

 number is increased by the mitosis of the tetracyte nucleus. 



In the mutual relationship of the Ustilaginaceae and Tilletiaceae, we 

 rely as much on pure speculation as on the interpretation of the pro- 

 mycelia. Although the Ustilaginaceous basidium is undoubtedly con- 

 nected to the Auricularia type, the terminal insertion of the sporidia on 

 the Tilletiaceous basidium points to an autobasidium. Hence it is not 

 impossible that both families, in spite of the similar structure of their 

 spores and the tendency to the formation of spore balls, in biological 

 relationship and pathological picture, represent two entirely distinct 

 phylogenetic lines, markedly convergent because of their parisitism. 

 When therefore we considered the Ustilaginales among the Phragmo- 

 basidiomycetes in connection with the Uredinales, this is only justified 

 for the Ustilaginaceae while for the Tilletiaceae there are many 

 possibilities. 



The ancestry of the Ustilaginales is not to be sought in the highly 

 specialized, parasitic, degenerate forms on gramineous hosts but in 

 genera of the Tuburcinia type. Vuillemin (1897, 1905) suggests the 

 Hypostomaceae whose two species, Meria Laricis and Hypostomum 

 Flichianum, greatly resemble the Ustilaginaceae, but, owing to lack of 

 cytological investigation, are not to be formulated in detail. Following 

 this suggestion, let us consider Septobasidium albidum, comparing the 

 conidial and basidial portions of its life cycle with the conidial and smut 

 spore portion of the Tuburcinia type. The roots of the Ustilaginaceae 

 would then (as those of the Uredinales with which they correspond in the 

 strong, already fixed, shifting of their developmental rhythm, eu-, endo- 

 and micro- type corresponding to Ustilago Zeae and U. violacea) be sought 



