July 29, 1920] 



NATURE 



681 



The Earliest Known Land Flora.^ 



By Prof. F. O 



'T'HE vegetable kingdom is made up of plants 

 J- of most varied size, character, and habitat. 

 Comparing those various types, the view becomes 

 ever more insistent that dependence on water is 

 the master-factor determining their existence. As 

 we range their diverse forms according to prob- 

 able sequences of descent, those which we regard 

 as the most primitive according to their structure 

 and mode of reproduction are those which are 

 habitually the most dependent upon constant 

 water supply. It is the same with the animal 

 kingdom. These broad results were summed up 

 by Weismann some forty years ago in the state- 

 ment that the birth-place of all animal and plant 

 life lay in the sea. If this be true, it follows that 

 all life on exposed land-surfaces has been second- 

 ary, and derivative. 



Geologists tell us that from the remote past 

 land-surfaces have stood exposed above the level 

 of the ocean. The continents and islands may 

 have differed from time to time in their outline 

 and area from those of the present day. But we 

 may believe that from a very early period land- 

 surfaces have had a continuous existence, so that 

 life upon land may itself have been continuous 

 from the time when living organisms first emerged 

 from their natal waters. Such beliefs throw back 

 to the very remote past the possible origin of 

 life upon dry land. But still the probability re- 

 mains that aquatic life antedated that event. These 

 considerations lead inevitably to the questions : 

 When was dry land first invaded from the water? 

 What were the first land-living plants and animals 

 like? And how did they rank as compared with 

 modern life? 



Leaving zoologists to solve these questions for 

 their own branch, we botanists are to-day in a 

 better position than ever before to answer thejn 

 with regard to plants. Though still far from being 

 able to visualise the beginning of the story, recent 

 discoveries have made it possible to see clearly and 

 in detail the nature of the earliest known land 

 flora, which is that of a period older than the 

 Upper Devonian. During recent years fossil 

 plants of early Devonian age have been 

 found in Sweden and in Scotland in greater 

 profusion than ever before, while the Scot- 

 tish specimens are so well preserved that 

 they are now almost as well known in structural 

 detail as plants of the present day. Already in 

 this room repeated lectures have been given on 

 the Palaeozoic flora. Many plants of the Carboni- 

 ferous Period have been described here in micro- 

 scopic detail, and they are mostly referable to 

 affinity with such living types as ferns, club-mosses 

 and horsetails. Some, such as the Sphenophylls 

 and Pteridosperms, represent classes which have 

 since died out. But, speaking generally, the flora 

 of the coal is composed of plants comparable with 



1 Discourse delivered at the Royal Institution on Friday, April 30. 



NO. 2648, VOL. I05I 



Bower, F.R.S. 



the lower vascular plants now living. They pos- 

 sessed stems, leaves, roots, and sporangia. Some 

 even produced seeds like modern Gymnosperms. 



Passing back from the Carboniferous period to 

 the Upper Devonian, the flora, though more re- 

 stricted, may still be described in terms applicable 

 to the living vegetation. The plants include 

 among others the gigantic fern-like Archaeopteris 

 hibernica, from Kiltorkan, co. Kilkenny ; the large 

 Lycopod Bothrodendron, from the same source ; 

 and the large-leaved Pseudobornia, from Bear 

 Island. Flat leaf-expansions are here seen, and 

 the plants named have been referred in their 

 general characters respectively to affinity with the 

 ferns, club-mosses, and horsetails. But between 

 the Upper Devonian and the strata that lie 

 below geologists tell us that a vast period 

 of time intervened. The evidence of the 

 ■plant-remains supports this. The earlier 

 Devonian fossils so far known are meagre 

 in number of forms. In their characters they 

 differ more markedly from the plants of the 

 present day than any of their successors. They 

 were rootless, and there appears to be a complete 

 absence of large, flattened leaf-expansions. It is 

 upon them that the new discoveries have shed so 

 interesting a light. Conversely, that light is re- 

 flected back by comparison upon the more recent 

 forms. In fact, a new chapter has been opened 

 in plant-morphology, and a new class of vascular 

 plants, the Psilophytales, has been established to 

 receive these representatives of the oldest known 

 land flora. The study of them is leading to new 

 interpretations of the form shown by plants of 

 later periods, and ultimately of the present day. 



Until 1913 the plants of the early Devonian 

 rocks were very imperfectly known. Their recog- 

 nised characters were chiefly negative. There was 

 no evidence of broad leaf-surfaces, nor was it 

 clear whether or not they bore leaves as distinct 

 from stems. The existence of true roots was also 

 doubtful. The best known plants were con- 

 structed of approximately cylindrical stalks bear- 

 ing lateral spines. These stalks arose from a 

 branched and creeping base. Some of them 

 showed crozier-like curves when young, and 

 sporangium-like bodies were sometimes found 

 upon them. The most distinctive of these plants 

 were grouped by Dawson in his genus Psilo- 

 phyton, and he published a reconstruction of the 

 species P. princeps. It was, however, the subject 

 of adverse criticism by his contemporaries, and 

 the validity of the genus was questioned. 



It was upon a field so open as this that light 

 has now been shed. From fresh-water deposits 

 of early Devonian age round Lake Roragen, 

 on the frontier between Norway and Sweden, Dr. 

 Halle collected many specimens of fossil plants. 

 But they were mostly impressions, and showed 

 only imperfect preservation of their microscopic 



