18 CONTRIBUTIONS TO WESTERN BOTANY. 



with age. This has hitherto been takai for E. gracile, which is 

 quite a different plant. It passes by insensible gradations into the 

 variety and is the most common of all the Great Basin species. E. 

 gracile has a prolonged and leafy stem and leaves oblanceolatc 

 and tapering into petiole. 



CHENOPODIACEAE, 



A recent study of this family emphasizes the remark of Wat- 

 son that the species and genera must be studied in the field in or- 

 der to obtain an accurate idea of their limitations. The work of 

 Watson was the best ever done on the family, and yet it is impos- 

 sible to accept his limitations or conclusions as a whole. The fam- 

 ily is remarkable for its insensitiveness to climatic influences, and 

 its consequent 'ndifference to zonal linitatons. The same species 

 will grow anywhere from the Middle Temperate to the Tropical 

 life zone, provided the soil is alkaline ; and in a few species, such 

 as IMouolepis ciicnopodioides, evni tiir soil need not i^-t alkaline 

 always. Considerable stress has been laid on leaf-character 



in classification, but this fails in most cases, particularly in Suaeda, 

 Chenopodium and Atriplex. In Atriplex the shape and append- 

 ages of the bracts have been made much of, but as will be shown 

 in this paper they are very uncertain quantities. Doubtless there 

 are twice as many species recognize<! as exist, but there is as yet 

 no way of determining the validity of half of the species except 

 by the most careful and long continued field work, which as yet 

 is a minus quantity. 



Suaeda Torreyana and S. intermedia do not hold their char- 

 acters as given by Watson. Both are said to be common in our 

 region, but the prevailing form has the leaves of S. intermedia 

 and the flowers of S. Torreyana. 



Suaeda occidentaJis Watson is supposed to have been found 



