2do 



Arthur: Nineteen Years of Culture Work 22 



with a more or less extended range of hosts in part determined 

 by cultures and in part by microscopical similarities in the fungus. 

 A species at the beginning of the work was conceived as a simple 

 and direct succession of individuals of the same appearance, 

 capable of being demonstrated by cultures, but at the close had 

 become a bundle of somewhat mobile characters, often compris- 

 ing many strains varying physiologically and sometimes morpho- 

 logically, and to a more or less extent not interchangeable by 

 cultures. 



In some other ways than already mentioned the accepted 

 notions regarding rusts were modified. It was found that telio- 

 spores among the grass forms were not all necessarily resting 

 spores, and that the non-resting forms presented special problems, 

 whose solution was not far advanced when the work came to n 

 close. Assistance with field observations and material permitted 

 successful cultures to be made in May, 191 1, with the aeciospores 

 from Arabis sown on Trisetum. The Arabis aecia arise from 

 systemic mycelium extending throughout the stem and leaves of 

 the plant. A month later teliospores resulting from this culture, 

 now having become mature, were found to be capable of germina- 

 tion and were sown on seedling rosettes of Arabis. The results 

 of this sowing first definitely showed when the axis of the Arabis 

 began to elongate as growth started the following spring. A 

 culture was similarly carried out in 1903 with Puccinia Eatoniac, 

 using the aecia on Ranunculus abortivus, also a form with dif- 

 fused mycelium, but a reciprocal culture was not made. These 

 two species of rusts, having a systemic form of aecia, were the 

 onlv ones of the kind which were brought under culture. They 

 belong to an interesting class physiologically, with systemic aecia, 

 and with teliospores capable of germination upon maturity, which 

 possibly do not retain their viability through the winter, or only 

 to an impaired degree. 



The culture work began with the too prevalent idea that alt 

 rusts could be expected to conform in general to the well known 

 Puccinia gram in is. It closed with the conviction that the rusts 

 are far too diversified in their morphology, their numerous char- 

 acters, their physiological adaptations, and their range of hosts, 



