THE AMERICAN BOTANIST 147 



plants which are regarded as among the early dicots. The 

 water lily until recently considered a member of the dicots in 

 good standing is now placed by some with the monocots. It 

 may be recalled, too, that several of the grasses have the re- 

 mains of a second cotyledon. 



Death of Charles K. Dodge. — Michigan botany lost a 

 devoted student in the death of Charles K. Dodge at Port 

 Huron, Michigan, March 22, 1918. Mr. Dodge was born April 

 26, 1844, in Blackmail Township, Jackson County, Michigan, 

 and lived all his life in that State with the exception of two 

 years spent in the West and South. He was in turn, school 

 teacher, lawyer and United States Customs officer holding the 

 latter position till his death. His "chief publications were lists 

 of the plants of northern Michigan and Lambton County, 

 Ontario. 



Age of Endemic Species. — In many parts of the world 

 one conies upon plants that have an extremely limited range. 

 They may, perhaps, be confined to a single mountain range, a 

 river valley, or an island. Usuallv botanists have considered 

 such species, to which the term endemic is applied, to be the 

 relics of species which once had a much wider distribution. 

 Under this supposition, they would be considered to be the 

 oldest species in the flora. There are, to be sure, species that 

 are properly considered relics of a wider distribution, such as 

 the Hart's-tongue fern (Scalopendrium) , which is common in 

 Europe but found in only one or two places in North America. 

 Endemic species, however, always occur in a single region. 

 The idea that they are among the oldest species has been 

 recently challenged by J. C. Willis, who by means of his "age 

 and area hypothesis," attempts to show that they are the very 

 youngest. According to Willis, in any group of twenty closely 

 related species, those that are most widely distributed show 

 by this fact that they are the oldest and have had time to thus 

 extend their ranges. He suggests that when a genus has a 



