1094 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXI. 



other with thin walls. (Fig. 6). There are no gTaclation? between them. 

 This observation recalls a similar one recorded for Melampnorella aspidiottis 

 in Engler and Prantl's Naturlichen Pflanzenfamilien I-Teil I-Abteilung, 

 p. 45. It was at first believed that these really represented two distinct 

 kinds of Uredo spores and that the thicker walled spores were of the nature 

 of Amphispores, which, (according to Arthur quoted in McAlpine's " The 

 rusts of Australia") "are developed in arid or subarid regions, being pro- 

 vided with thickened walls to enable them to withstand unfavourable condi- 

 tions, just like a teleutospore." But further observations throw a consider- 

 able doubt on this matter. If the spores from leaves kept dry in the 

 herbarium for a month or two are exainined, the proportion of thick-walled 

 spores to thin-walled ones appreciably increases, suggesting that the thin- 

 walled spores may develop a thick wall later on under certain conditions. 

 Again after treatment with lactic acid, the distinction between thin-walled 

 and thick-walled spores practically disappeared, the thin walls swelling 

 almost to the size of the thick walls. The difference in the appearance of 

 the spores examined fresh is, however, so striking that it is considered 

 worth recording. 



The fresh spores germinate in distilled water, readily in the months of 

 January and February. The germination is poor earlier in the season. 

 The two kinds of spores germinate alike. (Fig. 7). More than one germ tube 

 are generally protruded but only one grows out finally, the other becoming 

 disorganised. The largest number of germ tubes observed is five, four being 

 the commonest number. The largest number of germ spores observed is six. 



The size of the spores varies from 25-29" X 19-25". 



No other spore form has been observed so far. An Aecidium was found 

 on some Brinjal plants growing in the neighbourhood of the Castor plants 

 on the Agricultural College Favm, Poona. This has been identified, provi- 

 sionally, with Aecidium habunr/uensis by Dr. Butler. A few inoculation 

 experiments carried on with the Aecidiospores of this gave negative results 

 and there is at present no clue to answer the question whether the rust on 

 ■Castor has any other spore forms or whether it is one of those rust fungi 

 whose life cycle is restricted to the Uredo stage. 

 Description of Figures. 

 (Figures 1 and 3 are free hand drawings. The rest are drawn 

 with Camera lucida.) 



Fig. 1. — Portion of rusted leaf of Ricinus communis. Natural size. 



Fig. 2. — Transverse section of leaf through a young sorus, showing 

 peridium x 860. 



Fig. 3. — A mature sorus x 120 about, showing ruptured epidermis and 

 peridium. 



Fig. 4. — A spore-bearing hypha separated by manipulation x 400. 



Fig. 5. — Uredo-spore-bearing hyphfe and paraphyses x 400. 



Fig. f). — A thin-walled and a thick-walled spore x 500. 



