24 INVESTIGATIONS OF RUSTS. 



and Hoi way, a and a second instance of the occurrence of this form is 

 described by them for Paccinia trifisaci Diet, and Holw. The entire 

 series of observations and experiments with this species made by the 

 writer will now be described. 



The first cultures of amphispores were made January 15, 1891, at 

 Manhattan, Kans. Both a water culture and one of a sterilized decoc- 

 tion of manure were employed, with no result, the chief cause of fail- 

 ure being probably that the experiment was too early in the season. 

 Afterwards numerous other trials were made with no better success. 

 In the meantime true uredospores were found on September 25, 1896, 

 at Manhattan, Kans. 



Finally a successful culture of the amphispores was made on March 

 8, 1897, at Washington, D. C. The germination was excellent. An 

 abundance of rather long germ tubes, not promycelia, was produced 

 in forty -two hours, but only one to each spore. These germ tubes, 

 unlike those of most uredospores, are quite colorless and clear. A -few 

 teleutospores were present, none of which germinated. The culture 

 was an ordinaiy water drop, but was made in a new form of culture 

 cell, constructed to special order and similar to the Van Tieghem 

 cell, except that the glass ring is quite thin and drawn out into an 

 open tube on each of two opposite sides, with the opening plugged 

 with cotton wool, thus admitting sterilized air. This construction may 

 or may not have increased the chances for germination. The extreme 

 weather conditions at the time, to which the specimens were first 

 exposed, probably contributed most to the success of the culture. 

 They were fastened to the roof of a near-by shed, and after several 

 days of warm sunshine were thoroughly soaked with rain, which was 

 followed by snow and then a severe freeze, soon after Avhich the cul- 

 ture was made. It is an interesting feature of the experiment that 

 the specimens were received from Dr. David Griffiths, then at Aber- 

 deen, S. Dak., and had been collected in September, 1896, and kept in 

 the herbarium until sent to Washington. It was unfortunate that 

 seedlings of the host were not available for making inoculations with 

 these perfectly viable spores. Such an experiment is yet to be made. 



At Manhattan, Kans., in 1897, the uredo was present in considerable 

 abundance from June until late in October in grass plats on the 

 Experiment Station farm. On July 8 inoculations were made with 

 the uredospores of this species and with ^Ecidlum cephalanthi and 

 jE. xanthoxyli on seedlings of Bouteloua racemma without result. 

 Further inoculations with uredospores on October 1 were successful, 

 rust spots appearing in twelve days (PL II, fig. 9), followed in nine 

 more days with one sorus of amphispores. The uredosori are yel- 

 lowish-brown and rather inconspicuous. On December 31 uredo- 

 spores could not be found in the grass plats at Manhattan. 



"Bui. Lab. Nat. Hist., State Univ. Iowa, Vol. V, p. 175. 



