GEOGRAPHICAL DISTRIBUTION 175 



only assume that these genera are the little changed 

 descendants of ancient, widely distributed types. 



Although making a special search for Liverworts on 

 Krakatoa in 1906, Campbell found no specimens, nor up 

 to that time had any other collector. Professor Treub, 

 of the Botanic Garden at Buitenzorg, Java, had reported 

 two species of mosses. "Inasmuch as Krakatoa is 

 within sight of Java and Sumatra, both of which have 

 an extremely rich hepatic flora, the absence of these 

 plants from the new flora of Krakatoa is, to say the least, 

 worthy of note." In a similar way Campbell argues 

 that the wide distribution of mosses (cosmopolitan in 

 the case of the genus Sphagnum), combined with the 

 inability of their reproductive bodies to withstand trans- 

 portation over great distances, indicates a great antiquity 

 for the group; and this inference is substantiated by the 

 meager but positive evidence of fossil remains. 



In a later discussion of the origin of the Hawaiian 

 flora, Campbell 1 notes that the filmy ferns, since they are 

 hydrophytic with a rain-forest habit, and are, therefore, 

 not suited to transportation over wide stretches of ocean, 

 must have existed in Hawaii since those islands were 

 connected with some mainland, now submerged. The 

 relatively shallow water between Hawaii and the 

 Australasian-Malaysian regions, as compared to the 

 great depths between Hawaii and North America, in- 

 dicate a former mainland connection to the west, and 

 this inference is further substantiated by the great pre- 

 ponderance of Australasian-Malaysian plants in Hawaii 

 over those represented in America. In this connection 



1 Campbell, D. H. The origin of the Hawaiian flora. Mem. Torrey 

 Bot. Club, 17: 90-96. June, 1918. 



