122 ARTHUR: IMPORTANCE OF THE SPERMOGONIUM 
that spermogonia are directly associated with the teleutospores 
sets the matter at rest. The same is true when they are associated 
with the uredo. But when they are associated with the aecidia, 
the case is somewhat different. Only in instances where teleuto- 
spores and aecidia arise from the same mycelium, and immediately 
adjoining, can it be told from a specimen in hand that the cycle is 
without uredospores. 
It should be mentioned, that it is not to be inferred that when 
spore-forms are suppressed, all trace of them is lost. On the 
contrary, when the uredo-form is suppressed, a few uredospores 
are often found intermixed with the teleutospores ; and when both 
aedicia and uredo are suppressed, a few peridial cells and uredo- 
spores may occur in the teleutosori, reminders of two lost conidial 
generations. 
My second statement regarding the taxonomic value of the 
spermogonium is, that position, form, color and size, are characters 
worth recording, but have only minor value, as the range of vari- 
ation is slight. The spermogonium is usually a flask-shaped body, 
averaging 100-150 in diameter, formed immediately beneath the 
epidermis. The narrow neck protrudes through the epidermis, in 
order to discharge the spermatia into the air. Sometimes, especi- 
ally in certain species of rusts parasitic upon Ranunculaceae, Ana- 
cardiaceae, Rosaceae and Cuesalpiniaceae, the spermogonium is 
formed in the upper epidermal wall, just beneath the cuticle, and 
_ is then more or less hemispherical or conical. Intermediate forms 
sometimes occur, but the variation is usually inconsiderable. The 
position of the spermogonia in relation to the accompanying spore- 
form is also worth noting. The two structures are usually upon 
opposite sides of the leaf bearing them, but occasionally both are 
on the same side. The spermogonia sometimes occupy a small 
limited area, with the conidiosori on the opposite side of the leaf, 
or encircling them on the same side; or they may be scattered 
thickly or sparsely over large areas intermixed with the conidio- 
sori, or opposite them, 
The purpose of the paper has now beet attained, if the two- 
fold value of including information about the spermogonium in 
every descriptive account of a rust has been made clear. Spe! 
mogonial characters probably have little or no generic value, but 
