258 A TEXTBOOK OF THEORETICAL BOTANY 



Sexual Reproduction 



Initially there is a conjugation between short branches of two adjacent 

 hyphae, with the result that a fusion cell is formed (Fig. 250). There is no 

 evidence that any nuclear fusion occurs, though the nucleus of one hypha 

 passes into the other. This stage is followed by formation of additional cells, 

 which are budded off the fusion cell, or central cell. The cross walls of these 

 cells are not closed, and the whole structure forms a coenocyte or maze, as it 

 is sometimes called. Meanwhile, additional tissue grows up surrounding this 

 maze and forming the inner and possibly the outer wall of the perithecium. 



After a while the cells of the maze break down to form a cavity in 

 the centre of the perithecium, while by the formation of additional septa 

 the cells of the perithecium, which were originally multinucleate, become 

 progressively uninucleate. 



From near the base of the perithecium there then grow up into the cavity, 

 several cells each of which is binucleate. Their nuclei fuse, and in this way 

 a number of asci are produced. The fusion nuclei divide meiotically, followed 

 by a second and third division to form the eight nuclei of the ascospores, 

 around each of which the spore wall is ultimately formed. 



When mature the whole perithecium is blown away by the wind, and 

 may become attached to a fresh leaf by the appendages (Fig. 251). Here the 

 wall of the perithecium breaks down and the asci burst, liberating the 

 ascospores, which germinate, forming fresh hyphae. It is generally found that 

 the ascospores are not formed until the spring, since the perithecium over- 

 winters on the ground among dead grass, development stopping in the 

 autumn at the formation of the asci (Fig. 252). 



Biologic Races 



Recent work has shown that in Erysiphe graminis there are several biologic 

 races, that is to say that different strains exist, which are morphologically 

 indistinguishable but among which it can be shown that the strain which 

 attacks one species of grass will not infect any other grass. Similarly, 

 cultivated cereals each have their own particular strain. A point of some 

 interest is the existence of bridging species among the host plants of the 

 biologic strains of £". graminis. Thus the strain which infects Bromiis racemosus 

 will not attack B. commiitatiis directly, but it will attack B. hordeaceus, and 

 an infection from the latter species will attack B. commutatus. No explanation 

 is forthcoming as to what advantage the fungus derives from this very 

 specialized restriction of possible host plants. We shall, however, meet 

 with the same specialization in Piiccinia graminis. 



Similar biologic races occur in the types we have described. In 

 Sphaerotheca humuli there is a biologic form which occurs on the cultivated 

 strawberry, while in the United States there is a form which attacks roses. 

 In Erysiphe polygoni the form which attacks Peas will not affect either Swedes 

 or Turnips, while only the race on the latter plant is able to attack other 

 species of Brassica, especially Kale. In England, however, this latter form 

 rarely produces perithecia. 



