A FERN FORMING A XKW GENVS — ETHERmGE. 137 



The fourth example (PI. xxvi., fig. 8) is supplementary in some 

 points to PI. XXV., fig. 5, and is the longest portion of caudex in 

 the collection, but much decorticated. At one end are portions of 

 two fronds extended in opposite directions, and somewhat more 

 than halfway down are traces of two others, one attached, the other 

 protruding through the matrix, and although not attached, as in 

 the first instance, so clearly answering in position to the corres- 

 ponding frond above, on the right hand side, as to leave little 

 doubt that it also is in situ. 



Finally PI. xxiv., fig. 9, is the enlarged caudex of PI. xxiv., fig. 2, 

 and displays the broken edges of the different zones of the stem 

 which will be explained later on. 



The structure of the foregoing specimens may be summarised as 

 follows : — 



The Caudex. — The caudex is round, varying from one to three 

 millimetres in diameter, and in length from ten to forty-three 

 millimetres, so far as preserved, sometimes in the round, at other 

 times only as impressions, or both conditions may occur on the 

 same example. When in the former state there is clear evidence 

 of a peeling-ofF of layers, thus reducing the general bulk of the 

 caudex from what it must have really been in nature. At varied 

 and inequidistant points may be seen the minute thorn-like pro- 

 jections, when a caudex is seen in the round, or, in the case of an 

 impression, as small depressions. I am unable to offer any definite 

 explanation of these, but similar projections have been figured by 

 Mr. R. Kidston on problematical stems from the Lanarkshire 

 Ooal-field, called Psilotites unilateralis,* but as the name implies 

 they are on one side only, nor do I, by calling attention to the 

 resemblance mean to suggest any relation between the two plants. 

 Mr. A. C. Seward has suggested that these may mark the positions 

 of roots given off" from a creeping rhizome. 



The mode of distribution of the fronds on the caudex is peculiar, 

 and, so far as the specimens permit me to judge, constant. At 

 intervals occur clusters or tufts of fronds, the intervening caudex 

 surface being destitute of leaf clothing. A caudex is therefore 

 divided into nodes and internodes (PI. xxv., fig. 5 ; PL xxvi., 

 fig. 8). In PI. XXV., fig. 5, we observe a cluster proceeding from 

 an enlargement or corona, and at the other end a displaced cluster 

 that has been accidentally pressed backwards. In PI. xxvi., 

 fig. 8, traces of two of these nodes are visible, one a little below 

 the middle of the caudex, the second at the upper end. The 

 enlargement caused by the attachment of the frond bases has 

 almost weathered away, but to the right and left of each the edge 

 of a frond is traceable, particularly on the right of the lower tuft, 

 where it is distinguishable by its revolute margin, and traces of 



* Ann. Mag. Nat. Hist., xvii. (5), 1886, p. 495. 



