4G8 Scientific 'Proceedings, Royal Diihlin Society. 



induced to develop by cutting the leaf at right-angles in such a way as to 

 sever the main nerve. Again, by the exercise of no little gardening skill, 

 isolated leaves maybe rooted, the buds on the edge develop. But now comes 

 the difficulty. Invariably, as soon as the buds have got beyond the bud rudi- 

 ment stage, they begin to develop roots. Remove tliese roots laboriously day 

 by day, and for a few days tlie young sprout will grow, but always drops off 

 before it has reached even a few centimetres in height. It is possible that 

 methods might be devised to overcome these difficulties, but unfortunately 

 the time could not be spared to deal with them. The same difficulties crop up 

 in the others, save perhaps Begonia and Torenia, and they have been used 

 before. Perhaps something could be done with that Solanum Lycopersiciim 

 mentioned by Liitz (7), which developed buds in the axils of the secondary 

 leaves, which buds developed freely on plants whose main stem aud branches 

 had been removed. But I do not know the variety he used, nor did I ever meet 

 with a specimen of any of the ordinary Lycopersicums with such adventitious 

 buds. 



For these reasons, tlien, it was decided to make use of the first method, 

 namely, the grafting of sprouts on petioles. Here we can try any plant that 

 can be grafted, and this gives us a wide choice. We can make use of the most 

 ordinary plants of the greenhouse or garden, and thus silence tiie critic 

 referred to above. Thus it is a poor botanical garden, indeed, that cannot 

 supply the experimenter with Pelargoniums, aud the more detailed portion 

 of the work was carried out on one such — Pelargonium zonale v. meteor — the 

 despised geranium of the suburban window. The technique is simple, and 

 results can be obtained with the minimum of delay. 



II. — Material and Methods. 



Material. — The plants used were : — 



1. Pelargonium zonale v. meteor. 



2. Solamtm Riefiardi. 



3. Solatium Balhesii. 



4. Sanchezia nobilis. 



5. Phytolacca dioica. 



Not all of these were investigated as closely as meteor. Even of the results 

 given by them only such will be recorded as are necessary for the argument. 



Method. — The following methods were used : — 



A healthy plant was first chosen, and a large, well-developed leaf having 

 been decided upon, its lamina was removed by a transverse cut some little 

 way down the petiole. A large petiole was chosen, as the differences in size 



