CHAP, in.] of Cell-membrane in Plants. 271 



longitudinal extension of vesicles. He supported Bernhardi's 

 view of the nature of vessels, that the separable spiral threads 

 of spiral vessels are not wound round a membranous tube but 

 are surrounded by one. He maintains against Bernhardi the 

 distinctness of punctated vessels or porous woody tubes from 

 false tracheae or scalariform vessels, while he gave a more 

 correct description of the latter as they occur in Ferns. He 

 rejected Mirbel's view that the pits in dotted vessels are 

 depressions surrounded by a raised glandular edge, and ex- 

 plained them as grains or little spheres. Against this mistake 

 we may set off the very important step which he made in 

 advance, when he not only conjectured that the pitted vessels 

 of the wood are formed from cells previously divided off from 

 one another, but proved by observation that the members 

 composing such vessels are at first actually separated by 

 oblique cross-walls, which afterwards disappear. But this 

 correct observation was impaired by the mistaken idea, which 

 Treviranus shared with his predecessors, that the wood is the 

 result of transformation of the bast, and consequently that the 

 vessels of the wood are bast-fibres, which elongate considerably 

 after they are arranged in a direct chain one after the other ; 

 the unevennesses caused by the oblique junctions of the tissue 

 gradually disappear, the boundaries of each member of a vessel 

 being still for some time indicated by oblique transverse 

 markings. The dividing walls originally existing at these 

 points disappear by widening of the cavities, so that the 

 different parts come to form a continuous canal. To illustrate 

 the disappearance of a parting wall between two adjoining cells 

 Treviranus aptly points, somewhat to our surprise, to the 

 formation of the conjugating tube in Spirogyra. He rejects 

 with Bernhardi the view represented by Sprengel, Link, and 

 Rudolphi, that the different kinds of vessels are formed from 

 true spiral vessels ; he says that he had found the scalariform 

 ducts in Ferns so formed in their earliest stage and not as 

 spiral vessels; he thinks it highly probable that the distinct 



