TRICHOGYNE HYPOTHESIS 21 
From a comparison of all these observations it now becomes 
certain that, in the Uredinales, the typical mode by which the 
binucleate condition arises is by the conjugation of two similar 
cells, each provided with a large nucleus and an abundant 
supply of food. This fusion-cell can afterwards branch by the 
formation of lateral buds, usually in basipetal succession, and 
may thus produce several rows of spores at once ; by a similar 
branching bunches of uredospores and teleutospores can arise in 
the sori of those spore-forms (Blackman, ’06, Christman, ’07, 
Dittschlag, 10, Hoffmann, ’11). See Figs. 37, 156. 
At first, it may be supposed, the two conjugating cells . 
belonged to a definite basal layer or spore-bed, as in Phrag- 
midium, Puccinia spp., and Melampsora Rostrupii; but after- 
wards they ceased to be arranged in a layer, and conjugation 
took place between two purely vegetative cells in the mycelium 
beneath. The beginnings of this change are seen in I. 
Rostrupw, and its final product in such micro-forms as P. 
Adoxae where the greater part of the mycelium has synkarya. 
Whether the upper sterile cell which is so frequently met 
with is to be considered as an abortive trichogyne is not so 
certain. But it may be remarked that, since the sori generally 
arise beneath the epidermis, no fertilisation could have taken 
place by non-motile spermatia unless there were something of 
the nature of a trichogyne to protrude through a stoma. In 
this connection it is important to remember that in Uredinopsis, 
one of the lowest of the Uredinales, the sori of primary uredo- 
spores, i.e. ecidiospores, seem always to arise beneath a stoma ; 
other sori can arise in many genera in the same way, and in 
the genus Hemileia the pedicels of the uredospores protrude 
into the air through the stomatal pore. Moreover we know 
that, while ecidia of the enclosed higher types, as in P. Caricis, 
arise at some depth in the host-tissues, and the basal layer and 
its peridium are covered by a considerable thickness of dead 
empty cells (which are afterwards pushed to the sides), the 
eecidia of the more primitive form, the ceoma type, are shallow 
and are not enclosed in a peridium, but are either quite naked 
or surrounded only by a few paraphyses. 
In the typical Uredinales, the conjugation of the two 
