154 THE EVOLUTION OF PLANTS 



hand, the anatomical features of the Palaeozoic 

 allies of Marattiaceae are altogether those of 

 Ferns, and have none of the peculiarities of 

 Pteridosperm anatomy. On the whole, the prob- 

 ability is that a considerable proportion of the 

 supposed Palaeozoic Marattiaceae were really 

 Ferns of that group, though some have turned 

 out to be seed-bearing plants. The difficulty in 

 distinguishing between the two is in itself very 

 interesting, and suggests irresistibly that there 

 was a real affinity between Marattiaceae and 

 the early seed-bearing plants. Long before the 

 present evidence had been brought to light, 

 the late Professor Sachs used to teach that 

 Cycads were descended from Ferns of the family 

 Marattiaceae, and this view may prove to come 

 very near the truth. 



Assuming, as I think we safely may, that 

 Marattiaceae (or Marattiales, if we prefer that 

 more cautious termination, indicating a less 

 close relationship) really existed, as such, in 

 Palaeozoic times, it is evident that the Ferns of 

 that period were as highly developed as at the 

 present day, but very differently represented. 

 The Marattiaceous type, the highest of all, was 

 in full strength (the affinities with Pteridos- 

 perms are enough to prove the antiquity of the 

 group) and far more widely distributed than 

 now; other living families were for the most 



