228 



THE AMERICAN NATURALIST [Vol. XLl 



southern California. Its near of kin (without intermediate), E. 

 telmateia, is found with it (broadly speaking) in Europe and 

 California. Starting with E. pratense a next-related species is 

 E. sylvaticum. The former belongs to northern Europe, Siberia, 

 Alaska, Canada, the Rocky Mountains, Labrador, and south- 

 ward to Massachusetts and New Jersey. The other is circum- 

 boreal, covers the range of the first and with us goes somewhat 

 further south to Virginia. 



"In Isoetes we find the following coincident ranges of close 

 relatives. Isoetes tuckermanni is foiiiul (|iiite plentifully in New 

 England and completely overlies the range of its varieties harveyi 

 and horealis. I. Engelmanni is found plentifully throughout the 

 region east of the Appalachian range, from New Hampshire and 

 Vermont to Pennsylvania, extending sparingly to (Jeorgia. It 

 overlies the ranges of its varieties carol in iana, foiitaua and valida. 

 I. canadensis is found from Pennsylvania to ]Maine and Quebec, 

 appearing again in British Columbia. Its next of kin in the genus 

 is /. engelmanni, whose range for the most part it covers, and the 

 two species are not rarely found commingled in the same pond. 

 /. holandrri is found from Wyoming to (^ilifornia and Washing- 

 ton. Its next of kin would apjx'ar to be /. piigniwa of the Mono 

 Lake region of California, and the two .-.pecics were found by 

 members of the King Ivxpcdition in contiguous areas. It may 

 be supposed that /. puijnuni is an abnormal form of /. holandiTi 

 and hence not competent in this relation, but the next of kin of 

 bolanderi is /. echinospora var. hrainiii, which overlies the range 

 of bolanderi, but is widely distributed otherwiM- in North America." 



I have inquired of President lirainerd al)out the conditions in 

 Viola, and particularly whether pairs of closely related species 

 are found within the same range>. He an-wers: "Many ])airs 



