212 
composition, development, methods of reproduction and general 
behavior correspond to those of the Schizomycetes. 
Frank claims to have seen the spores of RAzzobium within the — 
Infektionsfaden. I have been unable to find anything in them 
except a granular, spongy (schaunmtges) appearance when treat- 
ed with K H O solution. These Infektionsfaden are not present 
in all tubercles, for instance, those of Phaseolus vulgaris, Pisum 
sativum, Robinia Pseudacacia. Nor are the “infecting strands” 
always absent where Frank thinks they are not needed. I have 
found them in young Phaseolus vulgaris; have traced them in 
tissues opposite the tubercle, or some distance from the tubercle, 
having no connection with it. Nor do these ‘‘strands” generally 
originate in a root hair, but more often from the epidermal layer 
of root cells, where they begin as a small spherical protoplasmic 
lump, as shown at a, fig. 1, plate CXXIX. These infecting strands 
seem to have no outer wall; they are nonseptate, branching, in- 
terlacing, passing directly through the protoplasm and wall. Their 
protoplasm is of a different consistency from that of either the 
cell or RAtzobia. It is more resistant toward chemicals and 
staining fluids. It remains after the protoplasm of cell and Rii- 
sobia have been used up. It is more than likely that these Infek- 
tionsfaden have nothing to do with the Rhizobia; because, they 
are not present in all cases, and when present, are not always in or 
near tubercles. I could detect no spores in them. They are 
present in tubercles of Zrifolium repens, whose Rhizobia have no 
spores. I found them in young Phaseolus vulgaris, where Frank 
says they are not needed. Since infection in some cases takes 
place without the “strands,” may it not do so in all cases? 
The form of the tubercle seem to have nothing to do with the 
kind of infecting Rhdzobia. In Dalea alopecuroides and Lathyrus 
odoratus, for instance, are found similar tubercles as to form and 
size, but they contain distinct Rhizobia. 1 did not notice much 
regularity as to the position and distribution of tubercles. They 
are usually more numerous near the surface of the ground. They 
may be found on the main root or on its branches. In 7rifolwm 
repens and T. pratense they seem to form more often where root- 
lets branch. The root hairs on tubercles are usually few and 
dwarfed, as shown in fig. 1, plate CKXIX. 
