44 PUCCINIA GRAMINIS 



origin of outbreaks <>f rust. It is uol inconsistent with this 

 thai Ward was able to prove that the mycelium of a uredo- 

 sorus extends only a little way round the margin of the sorus; 

 thai may be and is true in certain cases, especially with regard 

 to secondary uredo-sori, but in P.Caricis the mycelium extends 

 up and down the leaf between the parallel vascular bundles, 

 producing uredo-sori all along its course. The practical bearing 

 of Pritchard's discovery is to show that seed from an infected 

 crop should uever be used for planting. 



About the third possibility, it is difficult to come to any 

 definite conclusion. Eriksson's hypothesis is that the proto- 

 plasm of the fungus is present in the grain, mixed with the 

 protoplasm of the host, in such a way that the two are 

 indistinguishable. As the plant grows up, he supposes that 

 the two grow together until, at a certain time, the protoplasm 

 of the fungus separates itself from that of its host in the form 

 of " Nucleoli," passes into the intercellular spaces through 

 " invisible pores," then or earlier surrounds itself with a cell- 

 wall, forms a mycelium, and begins its ordinary life by producing 

 uredo-pustules. An intermediate stage, where the fungus- 

 protoplasm has surrounded itself by a cell-wall but is still 

 enclosed within the cells of its host, he named "special cor- 

 puscles." 



The difficulty in dealing with this theory lies in its in- 

 definiteness ; its author changes it from time to time to meet 

 objections, and supports it by hazy microscopical observations, 

 many of which are demonstrably the result of incorrect vision. 

 His " special corpuscles" have been shown by Ward and Klebahn 

 to be ordinary haustoria, Eriksson having completely overlooked 

 the intercellular hyphse to which those haustoria were attached. 

 It is incredible that the protoplasm of so highly evolved a fungus 

 could live outside its cell- walls, as he supposes. Such a state 

 of things is, of course, common in the lower fungi, Chytridina- 

 and allied groups. In Synchytrium Solum' the fungus-protoplasm 

 and the host-protoplasm may be seen in the same cell, before 

 the latter has been comjjletely devoured by the former, and in 

 that state they are even distinguishable by their microscopic- 

 appearance. But it will need a great deal more "proof" before 



