338 



^ 



U Mycologia 



L. umbellaium, with no infection on the first, and with abundant 

 infection on the latter, showing pycnia May 23 and aecia May 29. 

 Another sowing was made May 29 on Vagncra stcllata, Uvularia 

 grandiflora, Maianthemum canadense, Allium stellatum, all with 

 no infection, and also on A. Nuttallii giving pycnia June 4, and 

 aecia June 9, and on A. cernuum giving pycnia June 9, and aecia 

 June 12, both in abundance. Sowings June 5 on A. stellatum and 

 A. canadense gave no result. 



Although the host of this rust is one of the principal prairie 

 grasses of the western plains, yet the rust has been collected only 

 in three localities, one at Decorah, Iowa, in 1884, 1901 and 1902, 

 one at Ewing, northwestern Nebraska, in 1899, and the third a 

 recent find at Kulm, N. Dak. It differs from all other American 

 grass rusts in having the pores of the urediniospores near the 

 hilum. 



Last year Uromyccs Sporoboli E. & E. on Sporobolus vaginac- 

 fiorus was connected with aecia on Allium from Kansas material. 

 The aecia from the two species seem to be indistinguishable by 

 any technical character. The suggestion at once arises that the 

 two forms may be genetically correlated. Examination of the 

 telia supports this hypothesis, for the mesosporic teliospores of 

 the Puccinia are indistinguishable in size and appearance from the 

 teliopores of the Uromyces, but on the other hand the uredinio- 

 spores of the Puccinia are considerably smaller than those of the 

 Uromyccs and somewhat thinner walled. The greatest difference, 

 however, lies in the arrangement of pores of the urediniospores. 

 In the Puccinia they are three or four and basal, while in the 

 Uromyces they are four and equatorial. 



A particularly close correlation appears to exist between Puc- 

 cinia Cryptandri E. & B. on Sporobolus cryptandrus and U. Sporo- 

 boli, both in uredinia and telia, although the urediniospores of P. 

 Cryptandri are inclined to assume a winter resting condition with 

 thicker walls. Unfortunately no suitable culture of P. Cryptandri 

 has yet been made, and the aecia are unknown, but it may be 

 assumed that they will be found to occur on Allium or a closely 

 related host. 



Putting together what is now known regarding these three forms 

 of rust on Sporobolus and comparing with what is known of the 



