178 DIVISION II.—COURSE OF DEVELOPMENT OF FUNGI. 
The abjointed sporidia then unite in pairs in many species, either before or after 
their separation from the promycelium. They unite by means of short transverse 
processes, which may be placed at the point of insertion or at the apex or in the 
middle, and form double cells, as is shown in Figs. 81, 83, 84 3. Coalescence by 
pairs takes place also in the cases which come under No. 2 between the segment cells 
of the promycelium. Adjacent cells unite with one another by means of short 
processes arising near the partition-wall and forming with one another a short lateral 
excrescence (Fig. 84 C left), through which the protoplasm of the two cells coalesces 
after disappearance of the separating membrane. Communication is established 
between cells which are not next neighbours by larger lateral branches, which may 
form curved tubes or loops on a promycelium or even transverse bridges between two 
promycelia. 
The simplest form in which the development proceeds is by the production 
at a spot in the conjugated pair of a slender germ-fube with acropetal growth, which 
gradually receives the whole of the protoplasm of the two cells (Fig. 83 x); it has 
been shown that in many species this tube can penetrate into the phanerogamous 
plant which is its proper host, and there develope a mycelium which produces new 
resting-spores ; it may therefore be shortly described as an zuezpien? mycelium or 
mycelial primordium. 


FIG. 84. A Ustilago longissima, Tul, germinati BU. Tragopogonis, C U. Carbo, Tul. ~ promycelium, s primary 
poridia. Further ion in the text. .4 magn. nearly 700, 3 390, C more than 390 times, 

The process is more complicated in Tilletia, in many species of Entyloma, 
in Tuburcinia Trientalis, and as a general rule in Urocystis Violae, where a (secondary) 
sporidium is acrogenously abjointed on a short lateral branch from the conjugated pair 
(Figs. 83 s’, 81 d), which then gives rise to the incipient mycelium. 
Deviations from the above course of development occur under otherwise similar 
conditions in individuals of the species in which this development is the almost 
invariable rule. Firstly, the germ-tube which proceeds from the resting-spore may 
assume the characters not of a promycelium as described above, but of an incipient 
mycelium with acropetal growth but not producing sporidia. Secondly, the conjuga- 
tion of sporidia of the first order in pairs may be omitted, but the sporidia may still 
put forth tubes which are similar in conformation to the incipient mycelium ; this may 
happen to all or to a majority of the primary sporidia of a whorl in some species, for 
instance, of Entyloma, or, when the number of the members of a whorl is uneven, to 
one sporidium, while the others conjugate in pairs. 
These phenomena, which are individual exceptions in the cases which we have 
been hitherto considering, are the prevailing and even the invariable rule in another 
