August 5, 1920] 



NATURE 



m 



that of Psilotum, but the underground rhizomes 

 are longer and the appendages larger, while only 

 two loculi are usually present in each synangium. 

 Clearly the form and vascular structure of these 

 plants are generally like those of the Rhynie 

 flora. 



Until quite recently the Psilotaceae remained the 

 only living- Pteridophytes of which the life-cycle 

 was still incompletely known. In all the other 

 groups the regular alternation of two generations 

 had been demonstrated ; one is the prothallus, 

 which is sexual, and the other the established 

 plant, which is non-sexual. In the Psilotaceae also 

 the plant as above described is the non-sexual 

 generation, but hitherto the form or even the exist- 

 ence of the sexual generation remained problem- 

 atical. Since 1914 the prothalli of both genera of 

 the Psilotaceae have been discovered, and their 

 structure has been demonstrated by Darnell- 

 Smith, Lawson, and Holloway ; and so the very 

 last of these life-histories has now been completed. 

 It turns out that the prothallus of the Psilotaceae 

 is similar in its general characters to those of 

 other archaic Pteridophytes, being colourless, and 

 living in humus by means of fungal nourishment. 

 In fact, these plants conform in their life-cycle to 

 what is seen in the Lycopods and in the primitive 

 Ferns. Analogy with the living Psilotaceae makes 

 it highly probable that these early Devonian plants 

 also showed alternation. Though this has not 

 been demonstrated for them, their preservation is 

 so perfect that even the delicate prothallus maj 

 yet be revealed as the reward of further search. 



The interest of the recent work on the modern 

 Psilotaceae centres not so much in the details of 

 the prothallus as in their embryology. It has been 

 shown by Holloway that the embryo of Tmesi- 

 pteris is rootless from the first. This suggests that 

 the rootlessness is primitive, and not the result of 

 reduction. Since the Devonian plants were 

 rootless also, it seems probable that this state was 

 characteristic of such early plants of the land. 

 Further, the existence of Sporogonites, and the 

 very moss-like structure of its sporangium, to- 

 gether with its similarity to the sporangia of 

 Rhynia and Hornea, seem to link up the latter 

 naturally with the Bryophytes, which are also 

 rootless. In fact, we see before us a flora of 

 rootless plants, which r.aises afresh the question 

 of the first establishment of the neutral genera- 

 tion as an independent, soil-growing organism. 

 It originates in every case within the tissue of 

 the sexual plant, and is at first dependent upon 

 it. This condition is seen in the embryo of 

 Tmesipteris, with details not unlike those of the 

 Anthoceroteae. How, then, did it first establish 

 itself independently upon the soil? 



This question was first raised long ago by Dr. 

 Treub, the brilliant director of the Botanic Gar- 

 dens at Buitenzorg. He suggested that in the 

 evolution of land-living plants a rootless phase 

 would naturally precede the full establishment of 

 the sporophyte in the soil. He saw this reflected 

 in the embryonic state of certain Lycopods, where 

 a parenchymatous tuber precedes the establish- 

 NO. 2649, VOL. 105I 



ment of the rooted plant. It is attached to the 

 soil by rhizoids, and contains a mycorhizic fungus. 

 This tuber Treub styled the "protocorm." He 

 regarded it as a general precursor of the estab- 

 lished leafy plant in descent. During the war 

 new examples of this protocorm-stage were 

 described by Holloway, which show the condition 

 in its most pronounced form. In Lycopodium 

 laterale it constitutes the whole plant-body for 

 the first season. It bears numerous protophylls, 

 and may even branch, and reproduce itself vege- 

 tatively. It is only later that the leafy shoot and 

 lastly the root are formed. The fact that Hornea 

 shows a similar tuberous swelling at the base of 

 the rootless plant, and retains it even in the adult 

 state, brings the added interest that a permanent 

 protocorm figures in the earliest known land flora. 

 Its antiquity is thus undoubted. But* the 

 Devonian plants do not all show it in a distended 

 form. The tuberous swelling is not conspicuous 

 in Rhynia or in Asteroxylon, and it is significant 

 that in the living Tmesipteris the rhizome is 

 cylindrical. These facts indicate that the dis- 

 tended protocorm is neither an obligatory nor a 

 constant feature. 



It will not be necessary to do more than refer 

 briefly to the controversy whether the appendages 

 of the Psilotaceae are truly leaves or branches. 

 The fact suffices that the question has been in 

 debate, and that similar questions arise in rela- 

 tion to these fossils of the Devonian period. In 

 them it is impossible to assign the name "leaf" to 

 any definite part in the full sense in which it is 

 used in the higher vascular plants. The diffi- 

 culties of their morphological analysis and their 

 rootlessness are in themselves evidence of the 

 primitive state of these fossils. We are, in fact, 

 in the presence of what evolutionists call "syn- 

 thetic types "■ — that is, such as link together 

 groups which have diverged. The early 

 Devonian plants and the Psilotaceae show us just 

 those forms which might have been anticipated 

 as a consequence of comparative study, and some 

 of their characters were actually forecast by Dr. 

 Treub. 



Though it may be difficult to place the parts of 

 these synthetic types in the categories of stem, 

 leaf, and root, as those terms are applied to more 

 advanced forms, still they will serve to illuminate 

 the probable origin of these parts. The rhizomes 

 of Asteroxylon suggest an origin of roots from 

 branched, 'leafless rhizomes. Its " leaves " sug- 

 gest a relation with the leaves of Lycopods; but 

 its most significant feature is the branch-system 

 ascribed to Asteroxylon, bearing the distal 

 sporangia, which is so like that already described 

 for the enigmatical Carboniferous fossil Stauro- 

 pteris. This comparison has already been 

 pointed out by Kidston and Lang. On the other 

 hand, approaching the question from the side of 

 the living Ferns, I indicated in 1917 that "the 

 distal and marginal position of a sorus, often 

 monangial, is prevalent among primitive Ferns, 

 and that more complex sori are referable in origin 

 I to it." Comparison of the distal sporangia of 



