4 o4 SCIENCE PROGRESS. 



of Catamites were probably developed in a manner precisely 

 similar to that in Equisetum. 



Some years ago Williamson figured and described several 

 sections of well-preserved axes, showing a distinct resem- 

 blance to calamitean stems, but differing, as regards the 

 structure of the wood, in the absence of carinal canals. 

 These he at first regarded as some form of Catamites, 1 but 

 subsequently transferred them to a new genus Astromyelon} 

 That the earlier opinion was correct has been since demon- 

 strated by the discovery of French silicified specimens, in 

 which an Astromyelon axis occurs in direct continuity with 

 a Catamites stem, and has been shown to be the adventitious 

 root of that genus. 3 In young examples of this root there 

 have been recognised alternate strands of xylem and 

 phloem, the tracheids showing a distinct centripetal 

 manner of development characteristic of root structures. 

 The secondary wood agrees closely with that of the 

 stem. A constant feature of the cortex is its lacunar 

 zone ; there appears to be evidence of the existence of a 

 double endodermis, also a protective epidermoidal layer 

 immediately internal to the epidermis. 4 



In a recent work by Renault, of which plates were 

 published some months ago, but unfortunately without any 

 accompanying text, there are several excellent figures of 

 these adventitious roots of Catamites, spoken of in the 

 explanation of the figures as "stolons" of Arthropitys. b 

 The numerous specimens of Catamites represented in this 

 volume lead us to look forward with particular interest to 

 the appearance of the descriptive text. 



Among the plants of the Coal Measures there are fre- 

 quently found long and narrow strobili made up of a central 

 axis bearing verticils of linear leaves. Many of these have 

 been classified by Weiss, and referred to different provisional 



1 Williamson (i), pi. xxv., fig. 16. 



2 Ibid. (3), p. 319. :i Renault. 



4 A full description of these roots may be looked for in a forthcoming 

 paper by Williamson and Scott. 



5 Renault, pi. lvi., etc. 



