1887.] in the Neighhourliood of Simla. 371 



measure about 70 x 22 ft, and all are beset with prominent rod-like 

 excrescences (Fig. 5, PL XV.). When the secidium is ripe, the peridium 

 bursts longitudinally in strips from summit to base. 



The cpcidiospores are round or oftener oval, measuring, when 

 just wetted, 28 6 x 24*6 ft on an average. They are pale brown in 

 colour and their surfaces are beset with minute tubercles (Fig. 6, PL 

 XV.). 



The spermogonia are formed only on the upper surface of the leaf, 

 and when ripe have a sticky clear fluid over them probably secreted by 

 the tuft of paraphyses, which, arising from the very base of the organ, 

 project considerably (about 50 to 60 /x,) from the mouths. They are 

 well sunken into the tissue of the leaf, their bases pushing down and dis- 

 integrating the palisade-cells. They remain covered over by the cuti- 

 cle until ripe, when the cuticle is raised and opened by a porous opening. 

 The formation of the spermogonia precedes that of the secidia by a 

 very considerable length of time. In the spermogonial stage, the hy- 

 pertrophy of the leaf tissue is not so great as it becomes later, the 

 thickness of the lamina being about 0'41 m.m. If sections of the leaf 

 through a spermogonium be stained with Spiller's purple, it will be 

 found that, while the sterigmata are coloured blue, the spermatia are 

 coloured brick-red. The spermatia are oval and measure 8 x 4 ft. 



Hemarhs. — From the nature of the dehiscence of the peridium, there 

 can be no doubt that this ^cidium is identical with or allied to Gymno- 

 sporangium clavariceforme, Jacq., though in some respects it is not anlike 

 G. ju7iiperinum^ L., especially in the colour of the spores. I made very 

 frequent and numerous attempts to reproduce this ^cidium with teleu- 

 tospores from a Grymnosporangium on Gupressus torulosa which is by no 

 means uncommon on the few trees which occur in Simla, but always 

 without result. Notwithstanding these negative results, I am still, 

 however, inclined to believe that this Gymnosporangium (the only one 

 I am acquainted with in these parts) is genetically related to the ^ci- 

 dium just described, and I attribute my negative results to the influence 

 of some unknown condition attending my experiments, although I have 

 varied the conditions in every conceivable manner. 



15. — MoNOSPORiDiuM Andeachnis, gen. et. sp. nov. 

 Andrachne cordifolia, MiilL Arg. 



During August, especially towards its latter end, an ^cidium 

 is not unfrequently met with on this host, which is common at eleva- 

 tions a little below that of Simla. It is not, however, a common 

 -^cidium. Only the leaves are attacked and on them circular 

 patches are formed by the fungus, red above with an irregular halo of 



