No. 377-] REVIEWS OF RECENT LITERATURE. 365 
to the effect that the corolla arises as a distinct whorl of organs 
which are formed later than the inner sepals and receive bundles 
which leave the axis at a higher point. In the same way the long 
stamens are shown to be a distinct whorl and in no sense appendages 
of the shorter ones. 
Chodat and Lendner agree in nearly all points with the view of 
Klein, and argue that the seemingly bicarpellary gyncecium of the 
normal Cruciferee is in reality due to the union of four carpellary 
members. This view is based chiefly upon the course of the bundles 
in certain anomalous three- or four-carpelled specimens of Cheiranthus 
cheiri L. It is scarcely necessary to say that a conclusion from 
these rather doubtful premises must be accepted with all due caution. 
BLR 
Zinsser on Root Tubercles of Leguminosæ. — In Jahrb. f. wiss. 
Bot., Bd. xxx, Heft 4, pp. 423-452, may be found an interesting 
paper by O. Zinsser on the root tubercles of the Leguminosæ. 
This paper contradicts some of the statements of Frank, Gonner- 
mann, Laurent, etc., especially the statement that the root tubercle 
organism occurs outside of the tubercles in various parts of the 
plant. This work was done in the Botanical Institute at Leipzig. 
The following are some of the more important statements : 
1. Seeds of all sorts of leguminous plants were washed in sterile 
water, soaked fifteen minutes in water containing mercuric chloride 
(1: 1000), washed again thoroughly in sterile water, planted in ster- 
ile earth, covered with cotton-plugged sterile bell jars, and watered 
with sterile water. The plants which grew from these seeds were 
under observation eight to twelve weeks, but in no case did any 
tubercles form on their roots. If, however, the contents of root 
tubercles of these same plants was added to the earth, tubercles 
developed on the roots in most cases in fourteen days. The author 
believes with Prazmowski that Dr. Frank’s diametrically opposite 
results were due to the fact that he did not succeed in freeing his 
seeds from adhering surface organisms. So far as could be detected, 
the sublimate treatment did not in any way injure the plants. 
2. Other aérial parts and roots destitute of tubercles were then 
tested in various ways for the occurrence of the germ: 
(a) Approved staining methods, e.g., carbol fuchsin, alkaline 
methylen blue, gentian violet in anilin water, etc., were used on 
sections, but in no case could bacteria be demonstrated in the 
tissues. 
