258 E. J, BUTLER. 



were five palmyras from five to seven feet in height (to the base of 

 the crown), two of which were in low land under shade and three in 

 a drier situation, unshaded. 



The inoculations were made by removing the outer leaf-sheaths 

 until a white clean sheath was exposed, and applying a portion of 

 the material to the surface. The surface of the sheath was not arti- 

 ficially sterihzed, but was probably sterile, as the sheaths fit so close- 

 ly over one another that there is ordinarily no cavity left between. 

 The instruments used were flamed. The method does not differ in 

 any way from that commonly adopted in dealing with obligate para- 

 sites such as the rusts and mildews, as in the absence of artificial 

 cultivation of the parasite it is not possible to secure rigorously ster- 

 ile conditions.^ After inoculation the last removed leaf-sheath 

 was replaced in position and firmly bound in place. The wrappings 

 were watered at intervals for two days to prevent premature drying 

 of the material. 



On the 4th day after inoculation one of the trees was examined. 

 A discoloured patch was visible at the point where the material had 

 been placed. A day later the infected area was marked off from the 

 rest of the sheath by a marginal line, inside which were several iso- 

 lated spots of a deeper brown than the rest. Two days later the other 

 trees were examined, and a depressed area, marked by brown spots of 

 different sizes and surrounded by a raised edge along which ran a 

 reddish-brown bounding line, was found to occur in all, at the place 

 where the inoculations were made. The outer sheaths that had been 

 replaced over the infected ones had also become attacked and from 

 this time onward the point inoculated could be detected, without 

 unfastening the wrappings, by means of the spots which penetrated 

 them. By the 8th day one sheath was found completely penetra- 

 ted, the depressed area being larger on the inside than on the outside. 

 This was seen by lifting the free edge of the sheath without removing 



1 The nearest approach to such conditions has been obtained by the late Professor 

 Marshall Ward working with grass rusts. The ingenious method adopted is described in the 

 Proceedings of the Royal Society, Vol. LXIX, No. 458, 1902, p. 451, but is obviously 

 inapplicable to plants of the size of those under consideration. 



