6 14 STRUCTURE AND LIFE HISTORIES 
A consideration of the fossil record, however, makes it 
difficult to accept this hypothesis. Not only do we find, 
in the fossil forms described above, sporophytes that do 
not bear the remotest resemblance to the moss-sporo- 
gonium, but fossil mosses and liverworts have never been 
positively identified in either the Palaeozoic or the Meso- 
zoic rocks, while the same rocks are rich in fossils of such 
advanced forms as the broad-leaved sporophytes of the 
Cycadofilicales and Cycadophytes. We must not, how- 
ever, hastily conclude, from this lack of evidence, that 
mosses and liverworts did not exist in those early ages. 
Quite likely they were present when the Paleozoic rocks 
were being deposited, though doubtless not represented by 
the same genera, or at least not by the same species, as 
are now living. 
627. Summary of Results. From what has been said, 
in this and in Chapter XXXVI, we recognize that the 
method of evolution is to be ascertained chiefly by experi- 
ment by studying living plants in action; but the 
course of evolution chiefly by the study of comparative 
morphology, with special attention to fossil forms. Other 
points are necessary to complete the history of the 
evolution of plants; the above paragraphs give only the 
barest outline of the problem, for the entire history is 
much too long and much too difficult to be treated here. 
To summarize; the facts now known have led some 
investigators to infer: 
1. The origin of Angiosperms from Cycadophyta (pro- 
angiosperms) . 
2. The origin of Cycadophyta from Cycadofilicales. 
3. The origin of Cycadofilicales from Primofilices. 
4. The origin of Filicales from Primofilices. 
