136 BOTANICAL GAZETTE [aucust 
of amylodextrin or degenerate proteids, and which I have elsewhere 
designated as sporoids. The etiology and function of these bodies 
still remains to be determined. Some of the rhizobia showed distinct 
traces of forking (Y-forms), but the majority were of the Indian club 
form, derived from Y-forms, while a few were of uniform width, 
f evidently derived from simple 
unbranched rod forms. 
() The rhizobia of beach vetch were 
\ i i x rather remarkable for their branching 
(fig. 2). They present the general 
OD aA ne, morphological characteristics of the 
i cf rhizobia of sweet clover, bur clover, 
= SS and other vetches. The branching, 
d \ & however, is more pronounced than m 
\ \) any other form of this type hitherto 
Fic. 1.—Rhizobia from the 
root nodules of beach clover (Tri- 
jolium heterodon Gray), showing 
extreme form variation of R. muta- LP 
bile, due to hyper-nutrition; same 
organisms with the so-called spo- 
roids. 
examined. The branching is 
dichotomous and may be either 
unipolar or bipolar. The highly ‘\ 
refractive sporoids are not present, : a 
nor have they ever been observed in 
rhizobia of this type. : 
_ It is highly probable that these } 
' ‘ the root 
two rhizobia (of beach vetch and Fic. 2.—Rhizobia iro ae 
beach clover) represent twoextreme nodules of beach vetch ae the 
natural form types of Rhizobium  ™ritimus Bigel.), : ay: muta- 
: : extreme branching form 0! 
mutabile(R.leguminosarum Frank). ite, due to hyper-growth. 
Assuming that the rod forms and 
simple Y-forms are the original normal types, we have , 
rhizobia of beach clover (and in other clovers) the ex gr obia of 
deviation, apparently due to hyper-nutrition; and in the rhaiz0™ 
in the 
form 
