30 PROCEEDINGS OF THE 



would dispute this, the warniug was not superfluous, for names 

 exercise an undue influence and we are still ajit to think that 

 something is attained when we have referred a fossil to Fecoptens 

 or Neuropteris. Outline is, of course, more deceptive even than 

 venation. " On the whole," he remarks, " it is probable that 

 the irregularity of outline and division, prevalent in recent ferns, 

 is the most fertile source of error in our investigations amongst 

 the fossil" (p. W^). 



Considering the part played by glands in some recent investiga- 

 tions, it is worth noticing that Hooker specially calls attention 

 to the value of the characters afforded by hairs, scales, and 

 glands in living ferns (p. 414). Hooker's critical observations 

 on the study of fossil ferns have by no means diminished, but 

 rather gained in weight, now we know that under the name 

 " ferns '* so many Carboniferous plants of widely different aftinity 

 have been included. 



Passing on to Sigillaria, Hooker gives an interesting account 

 of the occurrence of the stems in coal-mines, showing a con- 

 siderable practical knowledge of the subject. He discusses the 

 possibility that some of the Lepidodendrons may have been the 

 branches of Sigillana, and adds that there is no real distinction 

 between the two genera (p. 416). At that time all kinds of 

 ideas as to the affinities of Sigillaria were held by good 

 authorities. Hooker thought it worth while to discuss, though 

 of course only to reject, the reference of this group to Euphor- 

 biacea?. Cacti, and Palmse (p. 420). The opinion that they were 

 ferns had the most advocates, though already abandoned, on 

 good grounds, by Lindley and Hutton. Hooker is rather favour- 

 able to the idea of some affinity, or at least analogy, between 

 Sigillaria and ferns, and even argues for the probability that 

 the Sigillarias may have borne fern-fronds (p. 417). He would 

 not admit that anything positive Avas known at that time of the 

 folliage of SlgiUaria^ for he was inclined to refer the only species 

 in which the true leaves had then been observed (>S'. hpidodendri- 

 folia) to Lepidodendron. Apart from his too liberal concessions 

 to the fern-theory. Hooker shows sound judgment as regards 

 Sigillarian affinities, for he says : " That the iSigillaria; were allied 

 to Lycopodiacea) is evident, their tissues and scarring being very 

 like those oi Lepidodendron" (p. 421). 



He recognised the high value of Brongniart's admirable account 

 of the anatomy of his SigiUaria elegans (really S. Menardi), but 

 had doubts whether the plant was a true Su/illaria. It will be 

 remembered that Erongniart was led by his discovery of radially 

 seriated (secondary) wood in tSigiUaria to refer that family to 

 the " great division of Gymnospermous Dicotyledons." He still 

 recognised some affinity to the Lycopodiaceae, regarding the 

 Sigilliirias as coming betAveen Lycopodiacd'e and Cycadeae, but 

 nearer to the latter. Hooker's remarks on this point are most 

 judicious. " Assuming," he says, " the 8. elegans to be a true 



