142 A. Barclay — On a second Species of Uredine affecting, ^c. [No. 2, 



merits may be seen coursing tlirougli tlie whole of the tissue underlying 

 the discoloured parts of needles, excepting the central bundles, which are 

 efficiently protected by the endodermal sheaths. Their structure may 

 be most conveniently observed in the air- cavities below stomata, where 

 they may generally be seen (fig. 6) . The hymenium is flat, and the secidio- 

 spores are given off in single rows from short sterigmata. Each spore of 

 a series is separated from adjoining ones by an intercalary cell or lamella. 

 These intercalary cells are best seen between the upper spores of a series, 

 being there larger and rounder, whilst between the lower spores they are 

 thin and wedge-shaped (fig. 3). 



The gecidiospores are for the most part oval, mieasuring on an aver- 

 age 30 X 24 ft, but varying from 24 x 24 /x to 34 x 22 //,. The epispore 

 is covered with prominent deciduous spines, which sometimes become 

 detached in flakes. The whole thickness of the epispore is about 2 to 3 />t. 

 They contain orange yellow granular contents sometimes with large oil 

 globules (fig. 4). The spores were on several occasions placed in water 

 in a moist cell with a view to observing their mode of germination, 

 but they persistently refused to grow. I had no opportunity of discover- 

 ing whether they would germinate after a period of quiescence ; pro- 

 bably such a period is necessary for them before germination is possible. 



The spermagonia are likewise deeply set with their bases below the 

 hypodermal layer of cells (fig. 8) : they have no hairs on their sum.- 

 mits and measure about 0"353 m.m. in width by 0"287 m.m. in depth. 



I looked carefully and often on all the surrounding vegetation for 

 indications of a further cycle of development of this uredine, but could 

 find no uredo- or teleutospore formation which I thought could possibly 

 be related to it. It has evidently no relationship with the uredine occur- 

 ring on the same host described in my former paper, for in the forests in 

 which that species was found in great abundance this one never occur- 

 red. Although the individual shoots attacked by the former parasite are 

 far more seriously injured, the entire shoot being killed by it, yet as 

 regards the whole tree the fungus just described must be the more 

 harmful owing to the much larger number of needles attacked which 

 fall early leaving the tree in a very ragged condition. 



EXPLANATION OF THE PLATES. 

 Plate IV. 



Fig. 1. General appearance of an affected shoot showing that only the nee- 

 dles of the last season's growth have been attacked and that many- 

 needles among those attacked have eacaped entirely. About natural 

 size. 

 „ 2. Fragments of mycelial filaments, x 340. 



