110 TECHNICAL BULLETIN 4. 



do they arise simultaneously but usually in succession. None has been 

 observed, however, which showed more than eight branches on a promy- 

 celium (Fig. 1, S). These primary branches are 2 to 3// in diameter, usu- 

 ally somewhat undulating, with broadly rounded tips. They soon become 

 septate, and almost invariably a lateral secondary branch grows out from 

 the top of the cell just below each septum (Fig. 1, 0-S"). The angle of 

 divergence between the primary and secondary branch is very broad, 

 often approaching a right angle. This manner of branching is character- 

 istic of onion smut mycelium wherever it is found, and is a good diagnostic 

 character. By continued branching, a dense mass of mycelium is developed 

 about the spore, and it becomes increasingly diflacult to follow the course 

 of single hyphse. Fig. 1, T, represents the latest stage in which the sep- 

 arate branches could be followed. On agar plates the older cells lose their 

 denoO protoplasmic content, and only the more distant tip cells appear 

 to be alive. In onion decoction, as the hyphse become older they become 

 more constricted at the septa and the cells rounded in the middle until 

 they appear almost separated from each other, the hypha having somewhat 

 the appearance of a string of beads t,Fig. 1, U). Frequently in the older 

 mats of mycelium from the germinated spore it has been observed that 

 some of the hyphal tips are recurved in the form of croziers. They have, 

 however, never been seen to develop further, and it is impossible to say 

 whether this development has any relation -to development of spores, 

 which are never produced except inside the tissue of the host. In hun- 

 dreds of germination tests which have been made during three years in a 

 large number of media, no conidia have ever been observed on the pro- 

 mycelium or its branches or anywhere else throughout the development 

 of the organism. Sometimes the short lateral branches appear like conidia, 

 but continued observation soon convinces one that they are merely vege- 

 tative branches which will elongate apically like other branches unless the 

 supply of nutriment is exhausted. 



Comparison with the Germination Process in Other Species of Urocystis. 

 Let us now compare this process with the process of germination which 

 other investigators have described for other species of Urocystis. 



Urocystis occulta Wallr., causing the flag smut of rye, was apparently 

 the first species of this genus which was studied with respect to germina- 

 tion of spores, that process having been first observed and described by 

 Kuhn in 1858. It was later studied by Wolff (19), Brefeld (4), McAlpine 

 (11) and others. According to Brefeld a promyceUal tube of varying 

 length is first produced, and at its apex it branches verticillately into a 

 whorl of four to six branches. These branches increase in length by apical 

 growth, and they, as well as the promycelial tube, become progressively 

 septate, while the protoplasmic content of the older cells constantly dis- 

 appears and the only living cells are those at, and just back of, the growing 

 tips. The verticillate branches never produce conidia, but form mycelium 

 by continued growth. McAlpine considers the verticillate branches them- 



