EXPERIMENTS IN THE CULTIVATION OF ONCOCYCLUS IRISES. 325 



supplied lime and magnesia to soils which were wanting in both, and 

 that, if possible, in a form which would be easily accessible to the roots of 

 the Oncocycli. As is well known, the dolomitic limestone contains a high 

 percentage of both lime and magnesia, and had it been possible for me 

 to get it here, I would certainly have added it to my soil after having it 

 ground to make it fine. The dolomitic limestone not being obtainable 

 here unless at very great cost, I tried to get a marl containing as high a 

 percentage of lime as possible, but found that most of the marl used in 

 agriculture also contained a dangerously high percentage of phosplioric 

 acid, Avhich being almost entirely absent in the Armenian soil, 1 deemed 



Fig. 178. — Iris pabadoxa var. Choschab. [The Garden.) 



it unwise to add to my ground. x\mong many samples of marl one 

 was at last found which only contained very feeble traces of the 

 phosphoric acid, and this was consequently selected. To provide the 

 magnesia the Grecian magnesite, which is easily enough obtainable, was 

 added. The question now arose, how much of the lime-holding marl 

 and of the magnesite was to be added to my soil ; and this, of course, 

 experiments only would show. Ten beds, each eight yards long by one 

 in width, were set apart for experimenting with, and when the rhizomes 

 of the Oncocycli were planted in November 1899, on the first bed a dose 

 of five kilos of marl and two of magnesite was strewn between and 



