Christman—Development of the Primary Uredospore. 521 
things, taken together with the fact that, as is well known for 
this form at least, the spermagonia accompany the formation 
of these so-called primary uredospores, suggests a possible close 
morphological relationship of aecidiospores and uredospores in 
general. I have given no figures of the spermagonia, hut my 
sections show that they are similar in every way to those accom¬ 
panying ordinary aecidia, except that they are frequently ex¬ 
panded laterally almost to the extent of forming crust-like 
masses. The further fact that the stalk cell of the primary 
uredospores is plainly, as described above, an elongated inter¬ 
calary cell strengthens the above conclusion very materially. 
Richards’ account (4) of the development of the aecidium 
in Uromyces caladii shows that in this' form a large, well-nour¬ 
ished hyphal branch, the “carpogonial branch,” is present 
as an initial structure in the formation of each aecidium. 
This structure is interpreted to be a remnant of a former spore- 
producing organ, which may have developed from a fertilized 
egg. We have no evidence that such a fertilization as this oc¬ 
curs in any form at the present time, and it is quite possible 
that these large hyphae are purely vegetative. I have been un¬ 
able to confirm Richards’ observations on my own material of 
Uromyces caladii. Still, the study of the initial stages of 
aecidium formation is not easy, and I am not certain that such 
a carpogonial hypha may not be present. Certainly no such 
structure is present in pustules of the caeoma type, unless, per¬ 
haps, it is comparable to the mass of hyphae which accumulate 
beneath the epidermis just before the gametes are formed. The 
discovery of such a carpogone would facilitate very much our 
interpretation of the aecidium. If Richards’ account be cor¬ 
rect, the conclusion is easy that the old process of fertilization 
by trichogyne and spermatia has been lost, and the new fertili¬ 
zations described by Blackman and myself have come to take 
its place. 
In a more recent work, Blackman and Eraser (2) have in¬ 
vestigated a number of rusts, and find that nuclear migration, 
like that found in Phragmidium violaceum, occurs also in 
Uromyces poae Raben. and Puccinia poarum Niels. In Puccinia 
