548 
PROFESSOR H. MARSHALL WARD ON THE 
to multiply. The ferment is now over. The gemmules come to rest in the matrix, 
which they have first stimulated and then exhausted, and a large vacuole may he 
formed (figs. 4 and 12), in which lie the remains of the nucleus. It should be noted 
that at this stage the nucleus is often very bright and fatty-looking, and stains black 
in osmic acid. By this time the filaments which budded off the tiny yeast-like 
gemmules are so obscured by their progeny that they are almost certain to be over¬ 
looked, and special methods are necessary to detect them. They are there, however— 
at least, recognisable remains of them are (fig. 18). 
The tubercle, when all its cells have undergone the above changes, now passes into 
a state of rest : it is a mass of cells full of yeast-cells—gemmules, germs —so tiny* 
that they might well be, as they were, mistaken for Bacteria. The rotting of the 
root and tubercle liberates these into the soil, and an extended acquaintance with 
these “ germs ” and their numbers leads one to feel no surprise if they turn out to be 
the ubiquitous germs which it has been suggested must exist to account for the 
universality of the root tubercles. 
Speculation apart, however, I have the following facts to offer. I made several 
attempts last year to infect the roots of the Bean by laying pieces of the tubers on 
the young rootlets of water-cultures. The success was only partial and doubtful, and 
the results did not seem sufficiently satisfactory to be worth recording. At the end 
of the summer I collected a number of the Bean-roots which had tubercles on them, 
and dried them ; these I kept through the autumn and winter, and made further 
experiments in January, February, and March of this spring (1887). Several of the 
infected plants had already developed several typical tubercles in March and April, 
and indeed the preparations from which figs. 7-9 and 11-13 were made have been 
obtained from these artificially infected tubercles. I have since repeated these 
experiments with marked success. 
The most curious feature about the matter is, perhaps, the long iC period of 
incubation” (as the doctors would term it). I infected the roots by placing very thin 
sections of the dried tubercles on the young tap-root; but no signs of the tubercles 
were noticed before the root was five weeks or so old, and had developed an abundant 
outgrowth of lateral rootlets.! It was from a section of one of the younger of these 
tubercles that the preparation of the root-hair in fig. 13 came. It will be noticed that 
the hypha running down the root-hair starts from a minute bright dot; unless this 
dot is one of the above-named “ germs,” I do not know what it can be. I have now 
seen the root-hair thus containing a hypha starting from a mere bright point many 
times, and have several permanent preparations of such infected hairs with the hypha 
passing down the cavity and across into the root. It very commonly happens that the 
* Frank gives their measurement as about O'OOl mm. thick. Woronin gives length '0016 to ‘0028 mm. 
Many Bacteria are larger. 
t They were, probably, at least a week or ten days old, however, and I now know that this specimen 
(growing slowly in January and February) furnished late examples. 
