ATKINSON: THE GENUS ENDOGONE 5 



No cross walls have been observed, except in the progarnetes after copu- 

 lation. The nuclei are minute, very numerous and lie in the peri- 

 pheral granular cytoplasm. There is a nuclear membrane and a 

 large nucleolus (? karyosome). The fruit bodies are quite firmly 

 attached to the living sphagnum leaves, but the mycelium does not 

 appear to be parasitic, although short haustoria have been seen pene- 

 trating the cell. The hold-fast is a rather compact lattice-like layer 

 of mycelium forming a kind of "sole," very closely applied to the 

 surface of the leaves, from which here and there the short haustoria 

 arise. The fungus is probably nourished by organic and mineral 

 solutions carried by the sphagnum from the water of the humus 

 substrate in the capillary stream so well provided for in the peat 

 mosses. 



Conjugation of the progarnetes. — While there is a great resemblance 

 in the process of conjugation and in the formation of the resting zygote 

 of Endogone sphagnophila to the situation in E. lactiflua, the details 

 of the process are quite different in the two species. The progamete 

 branches He nearly or quite parallel. In a few cases where they have 

 been observed just prior to conjugation they do not appear to be 

 differentiated from ordinary stout vegetative branches, except that 

 the cytoplasm is more dense and abundant. They do not appear to 

 be enlarged or clavate. In fact many of the vegetative branches are 

 clavate and sometimes they are in pairs lying closely side by side, 

 but in no case have I been able to determine with certainty that such 

 branches are progarnetes. The progametes also appear to be un- 

 differentiated before conjugation. They conjugate by lateral contact 

 of their walls at the tip. Immediately after contact the progametes 

 begin to swell into a clavate or fusoid form, and the wall at the point 

 of contact is resorbed for some distance, thus forming a broad com- 

 municating area where the cytoplasm of the two merges. During the 

 enlargement one of the gametes frequently becomes larger than the 

 other. The cytoplasm is very dense and fills the distal portion of the 

 progametes, while in the proximal direction the cytoplasm is less 

 abundant and lies chiefly in a peripheral zone next the wall. A cross 

 wall is now laid down in each progamete a short distance behind the 

 broad communicating pore, separating the gametangia from the stalks 

 or suspensors. 



Formation of the resting spore or resting zygote. — At the time of 

 conjugation and resorption of the contact wall the conjugating game- 

 tangia resembles the same stage of conjugation in Eremascus fertilis. 

 The zygote is not formed by the enlargement of the copulating game- 

 tangia as in the majority of the Mucorales, but the young zygote 

 begins to grow at once in an apical direction. Sometimes the origin 



