REVIEW OF GONDAVANALAND DATA 195 



families entering by this route), one from Africa to South America — 

 that is, via Gondwanaland (13 families traveling this route) — and a third 

 from Patagonia to Australia and New Zealand. Of the forms which came 

 via Gondwanaland some seem to have originated in Asia, others in Eu- 

 rope, and still others in Africa. A¥hile this is one of the best arguments 

 for the land connection, when dealing with land snails I can not get out 

 of my mind the remarkable records of the long periods snails have been 

 able to withdraw into their shells and go without food or water. The 

 lines of migration correspond surprisingly well with the directions fol- 

 lowed by the oceanic currents which touch the South American coast, and 

 this group, above all others, lends itself to distribution by floating rafts 

 or even driftwood. 



Perhaps the oldest argument for Gondwanaland is based on the Glos- 

 sopteris flora, a group of Pteridosperms characterized by the fronds being 

 simple and undissected, like those of the Avalking fern. Not all of the 

 flora, however, has simple fronds, but some are at least once pinnate like 

 Neuropteris or our common polypody. This flora, in its full develop- 

 ment, appears in South America in the Permian and just overlying the 

 glacial deposits. Soon, however — that is, in the higher beds — the flora 

 alters its character by the reappearance of fronds of more dissected type, 

 like Pecopteris and Neuropteris. This same sequence of events is also 

 found in India, Africa, and Australia. The genera typical of the Glos- 

 sopteris flora are generally placed among the pteridosperms on account of 

 the lack of any spore fronds and, in case of Glossopteris, because of flnd- 

 iiig seed cases. Such a genus as Neuropteris seems to be the nearest 

 northern relative. 



To my mind, the most pertinent question in regard to these southern 

 floras is whether the cold climate of the glaciation would cause the shape 

 of the frond to be altered, say from a once pinnate type to an undissected 

 type. My feeling in the matter is strongly influenced by having recently 

 looked over a large collection of ferns of the genus Polypodium. My idea 

 of what a polypody was like was obtained from seeing them in the woods 

 of northern United States, and I got the idea of a frond much like Neu- 

 ropteris. I found, however, that had I lived in China I should have had 

 a picture of polypodies with simple undissected fronds (glossopteris-like), 

 for there all were of this sort. Had I lived in the West Indies I should 

 have thought of them as either simple or once pinnate. The same would 

 be true of India or Hawaii, but had I gotten my idea in Europe I should 

 have thought of the fronds as once pinnate or even finely dissected. The 

 fronds of this one genus range from finely dissected, through the type 

 familiar to us all, to forms as simple as any Glossopteris. If this genus 

 of ferns varies so in the shape of its fronds (the reason for which I do not 



