110 MR. F. W. KEEBLE ON 



Secondly, experiments were made thus : — after the hypocotyl had almost reached its 

 full length, the seedUng Lot^anthus lonieeroides was so placed that the free end of the 

 suctorial disc came in contact obliquely with tinfoil. The root-end grew out in the 

 direction of the foil and in that region only, as is shown in fig. 9, PI. XI. The 

 suctorial disc, then, possesses contact-irritability, in consequence of which the *' root- 

 end " grows out toward the touching body. 



In yet another case this contact-irritability was demonstrated by an experiment 

 primarily devised to determine whether contact played a part in calling forth the rapid 

 swelling of the conical head of the hypocotyl which takes place when it reaches a branch. 

 A piece of glass cover-slip was pushed flatly on to the free surface of the suctorial 

 disc. The natural resin held it in its place. On looking through the glass the outline 

 of the disc could be seen, and marks were made on the glass coinciding with the limits of 

 the disc. 



After seven days the sucker's edge still coincided with these marks, showing thereby 

 that contact with an inert body had not caused a swelling of the sucker ; but during 

 this time a white *' root " had grown out and projected about 2 mm. beyond the rim 

 of the suctorial disc. Now the head of a hypocotyl which has reached a branch 

 ** swells " very rapidly indeed in the course even of one or two days. Whereas, then, the 

 head increases rapidly when its host is propitious, i. e. when it can penetrate the host, 

 no such increase occurs when the head is pressing even vigorously against an inert body. 

 It seems possible, therefore, to distinguish between the swelling of the head, which 

 occurs only when the sucker reaches a host into which it can penetrate, and the out- 

 growth of the root, which is brought about (at least to some extent) by contact pure and 

 simple. 



In order to examine early stages of penetration into the host, seedlings were grown on 

 bottle-cork. In these cases entry was effected (woodcut 2), and the hypocotyls swelled 

 considerably. AVhen, therefore, the head of the hypocotyl comes in contact with any 

 body, a contact-effect is produced, by which the root-end of the head grows out — not 

 necessarily swelling — and its superficial papillate cells press firmly against the opposing 

 body. If, however, the opposing surface is impenetrable — tinfoil, glass, etc., — these 

 papillate cells cannot penetrate, but continue to pour out large quantities of a semi- 

 watery resin. If the host is a branch or piece of bottle-cork, the suctorial disc swells 

 and the papillate cells effect an entry, apparently each of them for itself, into the cells of 

 the host, and instead of continuing to be merely or mainly secretors of resin they 

 become richly protoplasmic. They differ, moreover, in form : those growing into the 

 bark or cork become cylindrical, with rounded ends (woodcut 2) ; those growing freely 

 are much longer and more irregular in outline, and have richly granulated cell- 

 contents. 



I would suggest, therefore, that so far as these papillate cells are concerned, their 

 entry is partly a matter of pressure, for the outgrowing root-end of the sucker does 

 produce pressure sufficient to indent a fairly resistant surface, and, as I observed in 

 another case, to actually split the surface of the bark of a branch into which the root 



