266 SCIENCE PROGRESS. 



this character. Both these defects, however, are limited to 

 the earliest papers of the series, while from first to last 

 Williamson's account of the actual anatomical facts was in 

 all essentials correct. 



It would take too long to follow in any detail the pro- 

 gress of our author's work on the fossil Lycopodiacese. No 

 less than eight of the Royal Society memoirs are devoted 

 wholly or in part to these plants, besides which there are three 

 other treatises, namely, an important controversial paper 

 which appeared in the Annates des Sciences Natureltes for 

 1882 (28), the great monograph on Stigmaria ficoides (27), 

 1887, and the last paper he ever wrote, that on the mode 

 of growth of the Lepidodendra, which only appeared this 

 year (26). The unrivalled wealth of material in William- 

 son's possession, with which he was constantly supplied by 

 a band of enthusiastic fellow-workers, enabled him to place 

 our knowledge of the structure of Lepidodendrea^ on an 

 exceedingly broad basis. The anatomy of ten different 

 forms is described in his memoirs, most of which are known 

 by their leaves to be certainly true Lepidodendra, or to 

 belong to the closely allied genus Lepidophloios. In six 

 out of the ten the occurrence of secondary thickening of the 

 vascular cylinder (in addition to an enormous development 

 of periderm) has been proved. The only form which 

 attains any considerable size without developing secondary 

 wood is L. Harcourtii, the species which originally came 

 into Brongniart's hands, and unfortunately led him to his 

 hasty generalisation. It is a fact of special interest that 

 Williamson's L. futiginosum, a form so similar to L. Har- 

 courtii that it was for many years described under the 

 same name, develops at a rather late stage a well-marked 

 secondary zone. 



Specimens of Sigiltaria showing structure are rare 

 compared with those of Lepidodendron ; Williamson, how- 

 ever, was able to show that there is no essential anatomical 

 difference between Lepidodendroid and Sigillarian stems. 

 This established a strong presumption that, as the Lepido- 

 dendra are undoubtedly Lycopodiaceous Cryptogams, the 

 same holds good for Sigiltaria. This presumption became 



