56 
C. A. LUDWIG AND C. C. REES 
found without any special arrangement in the sorus the situation was 
not considered unusual. This condition is common with uredinia and, 
coupled with a marked indistinctness of the structures at the base of 
the sorus, led naturally to the conclusion that the spores were pedicel- 
late and that the short pedicels were concealed in what appeared to 
be a mass of intertwining hyphae. Under this impression the descrip- 
tions and classification of the Pucciniastratae were published by 
Arthur^ in the North American Flora, work previous to that time 
having thrown no doubt on the correctness of the view. 
The first intimation that the accepted idea might be wrong came 
w^ith the announcement by Liro" that the urediniospores of Melamp- 
sorella Cerastii (Pers.) Wint. are borne in chains. This point was 
developed somewhat and the sorus of M. Cerastii figured the next 
year in a paper by Magnus.^ The work of these two investigators, 
therefore, raised the question as to whether or not the urediniospores 
in the genera with similar sori were not also catenulate. It became, 
consequently, a matter of some importance to study these rusts in a 
more careful way than is possible with free-hand sections. 
The writers began their work on October 3, 191 3, by fixing some 
sori of P. Agrimoniae from leaves of Agrimonia parviflora Sol. The 
material was found growing near Lafayette, Indiana, and was the 
only fresh material of the group available. Part of it was fixed in 
chromoacetic acid and part in Flemming's weaker solution. It was 
imbedded in paraffin in the usual way and stained with the triple 
stain, the chief aim being to study the morphology of the sorus rather 
than nuclear phenomena. 
From the material in hand it has been possible to make out that 
in Pucciniastrum Agrimoniae (Schw.) Tranz. the uredinium begins 
as a small aggregation of hyphae under the epidermis. Presently 
some of the hyphae become erect, thickened, and divided by cross 
walls into three or four cells each. The apical cells of the columns 
elongate considerably and the protoplasmic contents become less 
dense, as is shown by a tendency to stain less deeply than at first. 
They are evidently the first peridial cells to be differentiated and the 
other cells of the chains are spores. No intercalary cells were seen 
and no chains were observed having more than three or four spores to 
1 N. Amer. Fl. 7: 97, 105-117. 1907. 
2 Uredineae Fennicae Finlands Rostsvampar 490, 492, 1908. 
3 Ber. Deutsch. Bot. Ges. 27: 320-327. 1909. 
