THE RUSTS OF NOVA SCOTIA. FRASER. 327 



eulieteruredo ; a eu-autoecious, euautouredo ', an opsis^heteroeci- 

 ous, opsisheteruredo, and the other combinations in a similar 

 manner. 



As the terminology of the spore structures in this paper- 

 is that proposed by Arthur, his definitions of the terms are 

 quoted (Bot. Gax. 39:221. 1905): 



"The terms I have to propose apply to the sorus. By sorus 

 is meant the structure which arises from a single fertile hyphal 

 mass or hyinenium, either with or without a peridium, now 

 usually called spermagonium, aecidium, uredosorus, teleu- 

 tosorus and kindred names. A sintple sorus includes the 

 peridium and all true paraphyses whether peripheral or discal. 



"The new terms consist of four words, with their deriva- 

 tives, one for each of the four stages of uredineal fungi. For 

 the initial stage, usually designated by a cipher and called 

 spermagonium, pyciiidium, etc., I propose pycnium: derivatives 

 pycnial, pycniospores, etc. For the sorus of the first spore- 

 stage usually designated by the Roman numeral I, and called 

 aecidium, roestelia, peridermium, etc., T propose aecium: 

 derivatives aecial, aeciospore, etc. For the sorus of the 

 second spore-stage, usually designated by the Roman numeral 

 II, and called uredosorus, etc., I propose uredinium (uredo) : 

 derivatives uredinal, urediniospores or if preferred uredospore, 

 etc. For the sorus of the third spore-stage, usually designated 

 by the Roman numeral TIT, and called teleutosorus, I propose 

 tclinin : derivatives telial, teliospores, etc." 



Specilization. 



Species have been based largely on morphological 

 characters. Infection experiments have led to a new conception 

 of species, based on the ability of the fungi to infect hosts. 

 Thus it has been found that some rusts morphologically alike 

 show differences in infecting power. On the other hand it has 

 been found that rusts which show differences in the choice of 



