THE FORMATION OF HYPHAL FUSIONS 39 



tion of the aecidiospores and the development of the diploid 

 mycelium, clamp-connexions are unknown. 



A clamp-connexion in a diploid mycelium, as we have seen, may 

 be formed in a very few minutes, but it may persist and be active 

 for many days, weeks, or months. Let us now turn our attention 

 away from the clamp-connexion's mode of formation and endeavour 

 to find in its location and in its structure some explanation of its 

 function. 



Among the facts which may bear upon the problem of the 

 function of the clamp-connexion in the diploid mycelium of 

 Hymenomycetes and Gastromycetes are the following : (1) in both 

 haploid and diploid mycelia each septum has a small central open 

 pore through which protoplasm can flow from cell to cell ; (2) under 

 natural conditions in the open, the haplophase of a mycelium is in 

 general of short duration and soon passes into the diplophase ; and 

 (3) since, as a rule, wild fruit -bodies, as well as wild sclerotia and 

 mycelial cords which act as magazines and subsequently produce 

 fruit-bodies, are diploid, these structures are formed not by haploid 

 mycelia but by diploid mycelia, so that it is the clamp-bearing 

 diploid mycelium, and not the clamp-less haploid mycelium which 

 has the task of evacuating its protoplasm and despatching it either 

 directly to the fruit-body or to temporary storage organs. 



The flow of labile protoplasm, driven by vacuolar pressure, 

 through the hyphal pipes of a mycelium encounters a certain amount 

 of resistance at each septum ; for the central pore of each septum, 

 through which the protoplasm must pass, is relatively small (about 

 1 (x in diameter). If, therefore, we suppose that of two hyphae, 

 where other things are equal, one has single septa between adjacent 

 cells, as in a haploid mycelium, and the other has twin septa between 

 adjacent cells, as in a diploid mycelium, it is clear that the flow of 

 protoplasm would be faster from cell to cell in the second or diploid 

 type of hypha than in the first or haploid type of hypha. 



In view of the above discussion it may be concluded that, in the 

 Hymenomycetes and Gastromycetes, the clamp-connexion may be 

 regarded as a means for providing between any two adjacent cells 

 of a diploid mycelium two septa instead of one and, therefore, two 

 passage-ways for the streaming of the protoplasm instead of one. Thus, 



