97 
The Evolution of the Higher Uredinece. 
The Leptopucciniae are omitted: there is not a special group 
of that kind. Schroter defined it as the group in which there are 
teleutospores only which germinate as soon as mature. But this 
is not absolutely true; most, if not all, the Leptopuccinise possess 
spores of both kinds, some which germinate at once, and others 
which do so only after the winter’s rest, e.g., P. Veronicarum. We 
should speak of lepto-forms of the groups: e.g., Bndophyllum 
Sempervivi is a lepto-form. 
Since this classification depends upon the number of known 
spore-forms, it is evident that a species must be changed from one 
section to another as our knowledge of it increases, e.g., P. Pruni- 
spinosce was to Schroter a Hemipuccinia, now it belongs to the 
Hetereupucciniae, while P. cohaesa belongs to the Auteupucciniae 
and P. fusca to the Micropuccinias. Schroter’s classification does 
not indicate affinity in any degree; its sections cannot have the 
force of sub-genera. 
Another more recent attempt at classification is that of 
the brothers Sydow (23). They arrange the species according to 
their hosts, beginning with those on the Compositse and ending 
with those on the Gramineas. In passing, there should be noted 
the false implication that the Compositae are the most highly 
evolved order and the Gramineae the least so. Here also allied 
species are occasionally widely separated, e.g., P. fusca and 
P. Pruni-spinosce, though P. fusca and P. cohcesa are brought close 
together. Their method is excellent as an index, and is, indeed, 
employed by them on that ground; but it cannot be used in any 
other way, although it does in a large number of cases bring allied 
species together, simply because many of those parasitic upon 
allied genera or species have been evolved pari passu with the 
latter from a common ancestor respectively. 
A third attempt is that of Fischer (12), who takes the form and 
markings of the teleutospore as his basis, but combines it with the 
grouping of the parasites according to their teleutospore hosts, 
or with the biological distinctions of Schroter. This is a great 
advance, because it brings together allied species, not only in the 
case of P. fusca, P. suffusca, P. Thalictri, but in many others; even 
if the heteroecism of P. Pruni-spinosce had been known in time, 
however, it might still have been separated from its allies. 
A fourth attempt is that of Arthur (1), who bases his genera, 
e.g., Tranzschelia, upon the number of spore-forms known com¬ 
bined with other morphological characters. This is also an 
