November 4, 1922] 



NA TURE 



607 



a conclusion as established until we are satisfied that 

 the surprisingly simple organisation of the Rhyniacea? 

 was really primitive. 



We can scarcely feel sure that a certain amount of 

 reduction may not have already been undergone, even 

 by this early race of land-plants. The presence of 

 stomata of the familiar type proves that the plants 

 must long have been adapted to a sub-aerial life ; at 

 the same time the small number of these organs 

 suggests xerophytic modification, which is quite con- 

 sistent with a peat-habitat. In fact the habitat warns 

 us that the Rhynie plants may not have been quite 

 typical representatives of the flora of their time. A 

 comparison with the Saltwort (Salicornia) of our mud- 

 flats has even been suggested ! Such a plant, how- 

 ever, bears obvious marks of reduction which are 

 wanting in the Rhyniaceae. 



We can do no more than leave the question open. 

 Probably we are justified in accepting Rhynia and 

 Hornea as members of a relatively primitive race, even 

 though their excessive simplicity may have been in 

 some part due to the peculiar conditions under which 

 they had to live. 



The presence of Asteroxylon in the same beds might 

 perhaps be taken as an argument, against the theory of 

 reduction, for Asteroxylon was. comparatively speak- 

 ing, a highly organised plant. It is true it had no 

 roots, but the branched aerial stem was well clothed 

 with leaves, and had very much the habit of a Club- 

 Moss (Lycopodium). The anatomy of the stem was 

 also quite complex compared with that of the Rhyni- 

 aceae, though the rhizome was as simple as theirs, and, 

 oddly enough, bore no hairs. Asteroxylon was a 

 larger plant than the others, and had nothing specially 

 primitive in its external aspect. The leaves were 

 peculiar, however, in having a very imperfect vascular 

 supplv. for the strand which ran out from the central 

 stele towards each leaf stopped short in the leaf-base 

 and never entered the blade. This is one of three 

 points which suggest a certain degree of possibly 

 primitive simplicity, the other two being the absence 

 of differentiated roots and the structure of the water- 

 conducting elements (tracheides). Though the wood 

 is well developed and rather complex in form, having 

 a stellate transverse section, all the tracheides are of 

 one kind, namely, spiral. 



Unfortunately, our knowledge of Asteroxylon is not 

 quite so satisfactory as in the previous cases, for the 

 fructification has never been found in connexion with 

 the plant. There are peculiar naked branches closely 

 associated with a few of the specimens, and with these 

 branches, again, sporangia are found in association. 

 The sporangia are quite different from those of the 

 Rhyniaceffi, but recall the fructification of some of the 

 Carboniferous ferns. If we assume that the naked 

 branches and the sporangia belonged to the Asteroxylon 

 we get a very remarkable combination of characters, 

 as pointed out by Kidston and Lang. While the 

 anatomy and morphology of the vegetative organs are 

 suggestive of Psilotaceae (a small tropical and sub- 

 tropical family, of uncertain affinities) and Lycopods, 

 the supposed fertile branches and sporangia would link 

 the plant to the ferns. We cannot, however, lay- 

 much stress on this surprising synthesis of diverse 

 characters until the connexion of the parts has been 



NO. 2766, VOL. I IO] 



established. At present there is no evidence beyond 

 intimate association. 



Kidston and Lang are inclined to identify Aster- 

 oxylon with Thursophyton. a Middle Devonian plant, 

 of Club-Moss-like habit, hitherto only known from 

 impressions. The fructification assigned by certain 

 writers to species of Thursophyton is, however, of a 

 Lycopodiaceous character, and totally different from 

 that attributed to Asteroxylon. 



Asteroxylon is included, together with the Rhyni- 

 aceae, in the class Psilophytales, of which Sir William 

 Dawson's genus Psilophyton, established in the 'fifties 

 of the last century, is the type. There is now no doubt 

 that Dawson's account of Psilophyton, so long dis- 

 credited, was substantially correct. The plant had a 

 branched rhizome and a forked upright stem, more or 

 less spiny. The fructification consisted of long terminal 

 spore-sacs, much like those of Rhynia. The morpho- 

 logical nature of the spines is disputed ; they may be 

 interpreted as rudimentary leaves or as mere out- 

 growths, like those of Rhynia Gwynne-V aughani . A 

 general affinity between Dawson's plant and the 

 Rhynie fossils is evident, but the exact relations remain 

 doubtful. The late Dr. Arber regarded Psilophyton as 

 identical with Rhynia ; Kidston and Lang, on the 

 other hand, have decided to place it in the same family 

 with Asteroxylon. Neither view is established. 



It may be mentioned that a possible Psilophyton has 

 recently been recorded by Prof. Halle from the Silurian 

 (Lower Ludlow) of Gothland, an interesting discovery, 

 if confirmed. 



Many plants besides the four species of Vasculares 

 were found at Rhynie. The most interesting is a speci- 

 men of Nematophycus (or Nematophyton), a genus 

 hitherto generally regarded as belonging to the Algae ; 

 it has a complex structure of interwoven filaments, and 

 some of the species, from other localities, attained a 

 gigantic size. It is very remarkable that a plant with 

 the structure of a highly organised seaweed should occur 

 in a purely terrestrial flora like that of the Rhynie chert- 

 bed. It may suggest that the Alga of the period were 

 doing a little transmigration on their own account. 



A number of genera of the early Devonian flora 

 have been recorded as impressions, showing little or 

 nothing of the internal structure. Space forbids our 

 describing them here. Some are much like Psilophyton, 

 others resemble Club-mosses, while others again have a 

 curiously Alga-like habit. An excellent summary of 

 our knowledge of the Devonian plants generally will Lie 

 found in Dr. Arber's little volume. 4 



The earlier (Lower and Middle) Devonian flora was 

 for the most part characterised by comparatively 

 simple types of land plants ; in some cases, as we have 

 seen, their simplicity was extreme. Even then, how- 

 ever, there is evidence that very much higher forms 

 existed. Thus the fossil known as Palmopitys Milleri, 

 from the Middle Old Red Sandstone of Cromarty, was 

 described by its discoverer, Hugh Miller, as a " Conifer- 

 ous tree," and really has the structure of a well- 

 organised Gymnosperm. Miller himself fully realised 

 the importance of his discovery, which has scarcely 

 received the attention from botanists which it deserves. 5 

 (To be continued.) 



4 E. A. N. Arber, " Devonian Floras, a Study of the Origin of Cormo- 

 phyta." Cambridge University Press, rg^i. 



5 Hugh Miller, " Footprints of the Creator," edition of 1861, p. ior. 



