268 W. G. FARLOW. 



concavity, just where the archegonia are generally found. 

 But wherever such scalariform vessels were present there 

 were no traces whatever of archegonia to be found, although 

 antheridia were always abundant, as well as the hairs, which 

 here fulfil the offices of roots. (See PL X,figs. 1 and 2, in which 

 a shows the position of the scalariform ducts.) As may be 

 seen from figs. 6 and 9, the scalariform ducts arise singly, 

 and are situated in the central portion of the tissue of the 

 prothallus. They scarcely differ in shape at first from the 

 adjoining cells, which are longer and relatively narrower 

 than the superficial cells. The ducts increase by division in 

 a direction parallel to the surface, so that, in a longitudinal 

 section, we find several lying one above the other. 



Another peculiarity often, but not always, accompanying 

 the presence of scalariform ducts, was the formation of a 

 process or outgrowth in the concavity of the thallus, as shown 

 in PL X, figs. 1 and 2. This outgrowth was variable in length, 

 often being short and imbedded between the lateral lobes, but 

 sometimes projecting as a narrow tapering process. In one 

 case, it was forked at the extremity. The growth by means 

 of a single terminal cell is shown in PL XI, fig. 9. As just 

 mentioned, the existence of a process in the concavity is a 

 striking peculiarity, but not quite a constant occurrence, 

 like the presence of scalariform ducts. The first scalariform 

 duct arises in the prothallus, as I have just remarked; and 

 others soon appear, always in a line between the original 

 duct and the nearest point of the concavity. In this way arises 

 an interrujDted row of ducts, which may extend, when a 

 process is present, nearly to its extremity. The cells sur- 

 j-ounding the original duct soon assume the form of ducts 

 themselves, and thus a rudimentary bundle is formed. It 

 happens rarely that two such scalariform ducts apj)ear simul- 

 taneously in parts of the prothallus remote from one another. 

 I only saw one such case. 



It now becomes necessary to consider the relation of the 

 scalariform ducts to the other cells of the prothallus, and this 

 must be done by making longitudinal and transverse sections 

 of the region in which the ducts lie. From such longitudinal 

 sections (PL X, figs. 6 and 9), we see that the prothallus forms 

 a compact tissue in which certain cells have assumed the cha- 

 racter of scalariform ducts, while the others remain unchanged. 

 From no section made was I able to see any trace of an arche- 

 gonium. In two instances, when seen from above, a com- 

 bination of four cells led me to suppose that there was some 

 signification to be attached to this arrangement. But as a 

 longitudinal section (PL X, fig. 7), shows no connection 



