312 AFFINITIES [CH. 



authors it has been placed in the Haloragaceae 1 , but it is ex- 

 cluded by others, and a remote position near the Santalaceae 

 has even been assigned to it 2 . The most reasonable view seems 

 to be the non-committal one of Juel 3 , whose investigations led 

 him to believe that the position of the genus must still be treated 

 as uncertain, sipce it is by no means even proved that it belongs 

 to the Archichlamydeae. So it is best, provisionally, to relegate 

 it to a separate family, the Hippuridaceae, possibly allied to the 

 Haloragaceae. The geographical distribution of the two fami- 

 lies, as Schindler 2 has pointed out, lends colour to the idea of 

 their distinctness. He shows that the Haloragaceae (including 

 the two tribes Haloragideae and Gunnereae) form an "ant- 

 arctic" group of plants, a few of which by virtue of their 

 special dispersal-capacity as aquatics, extend into the north 

 temperate zone; while the Hippuridaceae, on the contrary, are 

 an "arctic" family, confined to the Northern Hemisphere. 



At different times in the last century, botanical writers have 

 grouped the following genera in pairs as members of the same 

 family Ceratophyllum with Callitriche, Callitriche with Myrio- 

 phyllum^ and Myriophyllum with Hippuris but more recent 

 research has led to the belief that these four genera may even 

 belong to four different Cohorts; this example indicates 

 the degree to which homoplastic convergence may prevail 

 among aquatics, and the confusion which it is apt to introduce 

 into systematic botany. 



The Umbelliferae are primarily terrestrial, but certain genera 

 and species have, to a greater or less degree, taken to aquatic 

 life. In some, e.g. Oenanthe Phellandrium^ var. fluviatilis, the 

 vegetative organs are completely submerged. 



Among the Sympetalae, water plants are more scattered, and 

 there is a notable absence of wholly hydrophytic families. 



In the Primulaceae there is the single aquatic genus, Hottonia, 

 with one European and one American representative. 



The Gentianaceae are mainly terrestrial, but such marsh 



1 Parmentier, P. (1897). 2 Schindler, A. K. (1904). 



3 Juel, O. (1910) and (191 1). 



