58 
C. A. LUDWIG AND C. C. REES 
of more nearly approximating a natural order. It has now been shown 
the plate-like layer is not very evident and in mature sori the chains 
are difficult to make out. 
The spores and other cells which go to make up the uredinia, and 
the telia as well, are binucleate. Such a condition is naturally ex- 
pected when the sporophytic stage of a pleomorphic rust is under 
consideration. The nuclear state of the free hyphae could not, how- 
ever, be determined ; but doubtless they are also binucleate. Nothing 
in the nature of hyphal fusions was observed. This is not unexpected 
in the consideration of the uredinial habits of this species, for normally 
fusions occur in the aecial stage; and all species of the genus Puc- 
ciniastrum are presumably heteroecious. 
The mature urediniospores (fig. 4) of Pucciniastrum Agrimoniae 
are obovoid in shape, not angular, and, as already mentioned, have a 
more or less distinct hilum. The wall is distinctly echinulate. These 
are characters commonly associated with pedicellate spores, and 
when found in catenulate spores, where only short chains can be seen, 
must evidently be taken to indicate that the terminal spore matures 
and becomes detached before the next one has advanced far in its 
development. The spore next to the free end of the chain then de- 
velops in exactly the same manner, and so the process is repeated as 
long as spore production continues. The association of the catenulate 
habit and the echinulate condition is worthy of remark. The writers 
are unfamiliar with any other genus, except Melampsorella, in which 
such association has been shown, although presumptively it may also 
occur in the nearly related Melampsoridium. Catenulate spores are 
usually verrucose except in the case of some teliospores, in which 
instance they are smooth, never echinulate. 
The peridium, as mentioned above, is formed in a way analogous 
to that in the ordinary aecidium. One wonders how it is produced in 
similar sori with pedicellate spores, as in the fern genera Uredinopsis, 
Hyalopsora, and Milesia, and could wish that Bartholomew's^ recent 
work with Hyalopsora Polypodii had cleared up the point for that 
species. It would perhaps not be too daring to risk the opinion that 
in such cases the roof cells are formed from the first urediniospores 
and produced in the same manner as in sori in which the spores are 
borne in chains. 
With the data now at hand it would seem to be possible to suggest a 
new arrangement of the genera of the Pucciniastratae with some hope 
- Bull. Torrey Bot. Club 43: 195-199. 1916. 
