THE ORIGIN OF GYMNOSPERMS 223 



but very recent, researches, which have proved so fruitful in 

 this respect. 



It has been known for nearly a century that seeds, similar 

 in organisation to those of modern Gymnosperms, occur in 

 rocks which are among the oldest in which fossil plants are 

 found. Thus the race of Gymnospermous seed-plants is geo- 

 logical^ ancient. Curiously enough, as will be seen at a later 

 stage in this consideration, one of the results of modern work 

 has been a tendency to push back, in the scale of geological 

 time, the appearance of the naked seed-bearing habit to an even 

 earlier period. This particular form of heterospory is found to 

 be incredibly old in a geological sense — far more ancient, in fact, 

 than those rocks in which the earliest known fossil plants are 

 found at present. Yet, while a primitive type of seed-plant is 

 still quite unknown to us in the fossil state, a study of the 

 oldest seed-plants with which we are acquainted has afforded 

 distinct evidence as to the ancestors from which some, at least, 

 of them were derived. 



The particular members of the Gymnospermous stock, whose 

 line of descent has now become clear, are the Cycads. We know 

 that the modern Cycads, 1 of the Tropics or sub-Tropics of both 

 the Old and New Worlds, are but survivals of what was once 

 a great group, holding a position in Mesozoic times similar 

 to that which the Dicotyledonous Flowering Plants hold in the 

 vegetation of to-day. This group has now been traced back 

 to a more ancient type of seed-bearer, long extinct, and this 

 in turn to a still older stock, which was homosporous, and 

 essentially Cryptogamic, as opposed to seed-bearing. 



A glance at the fossil plants of Carboniferous age preserved 

 in our larger museums, or even the casual collection of 

 specimens on the waste heaps of some colliery, will serve 

 to impress at least one fact. It will be found that impressions 

 of fronds — such as those known technically as Sphenopteris, 

 Neuroptcris, Alcthopteris, etc., all of which closely recall the 

 leaves of the modern ferns — were not only extremely 

 abundant, but greatly varied, in Carboniferous times. The 

 casual observer perhaps would be startled to learn that the 

 majority of such fossils, so very fern-like in every aspect, 

 were not ferns at all. Such, however, is the fact. Not only 



1 The present-day Cycads are included in about nine genera, and from seventy- 

 five to a hundred species. 



