544 
PROFESSOR H. MARSHALL WARD OH THE 
any case, again, it seems clear from further research that the infecting agent is 
present in the soil practically everywhere, and it is by no means difficult to suppose 
that it might attach itself to a seed harvested in the ordinary manner if the roots 
and soil are disturbed so that germs could be scattered. 
The tubercles are formed without any order on the tap-root and lateral roots, the 
only apparent rule being that they are not developed until the young root-system is 
fairly advanced,* and they do not arise close to the apex of the root: they seem to 
especially affect the region where the root-hairs are in full vigour. There may be only 
a few, or very many, crowded in groups or scattered (fig. 1). 
They are at first very small, and much like young rootlets in appearance; but they 
soon swell, often very irregularly, and may become lobed in various ways. Their 
colour and texture are quite like those of the rest of the root. 
A longitudinal section (figs. 2, 3, 7) through the tubercle shows the structures 
described by Eriksson. The axial chief mass of the tubercle consists of rather large 
polyhedral parenchymatous cells, passing at the apex into smaller, closely packed, thin- 
walled cells which constitute a meristem, which would be homologous with the growing- 
point of a rootlet : several layers of more compressed parenchymatous cells envelope 
the above tissues, and may be looked upon as a periblem (fig. 7). There is no root-cap. 
A short distance from the meristem, rows of cells at the boundary between this 
periblem and the axial (plerome) cylinder gradually pass over into vascular strands. 
Perhaps the outer layer of the periblem mantle may be regarded as the homologue of 
an epidermis and root-cap, but the resemblance is not very obvious. In the lobed or 
convoluted older tubercles (figs. 1, 2) all these tissues take part in forming the lobes. 
It is in the large-celled axial tissue that the parasite is rampant (fig. 3), and the 
cells which contain the densely crowded corpuscles (figs. 4, 5, 12) are seen on the 
section as slightly pinkish or bull-coloured masses sending ramifications into the 
various lobes (fig. 2). 
Any cell at the base of this mass may be seen to contain a densely granulated mass 
of substance (figs. 4, 5), which swells in water, and allows the corpuscles to escape 
passively, but with the well-known dancing Brownian movement. The separated 
corpuscles are very brilliant, and vary in size and shape; some are rod-shaped, others 
have the form of a Y or V, and others are still more branched, as described by Frank. 
(Fig. 6.) 
These corpuscles are deeply stained by hsematoxylin, and become yellow-brown in 
iodine ; in chlor-zinc iodine the mass of corpuscles turns bright golden-yellow, the cell- 
walls of the parenchyma containing them turn blue. All their reactions point to the 
accuracy of the previous ideas as to their nature; they are unquestionably organised 
bodies. 
I have repeatedly examined them in sections of old dried tubercles gathered the 
year before. If kept dry during the winter, the tubercle shrinks considerably, and 
* Usually wlien the seedling is about three to four weeks old, if growing luxuriantly. 
