TWENTY- SECOND ANNUAL MEETING. 33 



a few facts at present. Any additions to these facts, or any definite conclusions of 

 value that may be drawn from the facts, must be left for future papers. An intima- 

 tion here of what I have further in mind in regard to this subject might however 

 not be amiss. The idea has naturally enough suggested itself that these species 

 may possibly bear to each other a somewhat similar relation in regard to structure, 

 composition, and vitality, to that which they bear in regard to habitat. This is one 

 of the questions, then, among those that may be found worthy of careful considera- 

 tion. I am indebted to Prof. J. M. Coulter of Indiana and Prof. W. A. Kellerman 

 of the Kansas Agricultural College, for aid in identifying some of the species. 



SPECIES OBSEBVED. 



Priinus chicasa, Mx. — This is the most common sand-hill species. I have never 

 noted its absence on any sand-hill of ordinary extent, and it sometimes grows 

 where it seems nothing else can grow. It is found on some of the most recently 

 formed hills, near to rivers, as well as on the older hills. Very common over a great 

 portion of Sedgwick county, which is to a great extent bottom land, and therefore 

 sandy. In a few places only in Cloud and Saline counties, and always on sandy 

 wastes or rocky points. 



LithospeTmum hirtum, Lehm. — I have never observed this species, I think, out- 

 side of sandy districts, although I presume it has a more general distribution. Have 

 observed and collected it in several places in Sedgwick county, and in one spot only 

 in Cloud county. 



Evolvulus argenteus, Ph.— Have collected very few specimens of this species; 

 chiefly in Cloud county, on sandy wastes. It is reported from extreme southwest 

 Kansas, where the country is quite sandy and somewhat barren. 



Yucca angustifolia, Ph. — This species, while quite common, I think is not often 

 found in low, moist or rich soils. I have almost invariably found it on sandy wastes, 

 rocky hillsides, or in barren districts; and I have observed its habitat from Kansas 

 to Montana. It is very abundant in western Kansas and eastern Colorado, which we 

 all know to be quite a barren region. It is reported to be very abundant on the 

 "Staked Plains" of Texas, growing, it is said, to the height of ten feet — probably 

 a different species. The yucca soap manufacturers, of Wichita, obtain their ma- 

 terial from the Texas yucca. In Riley county I have collected it invariably on rocky 

 hillsides. On the extreme western plains it is associated with the sage-brush, 

 prickly-pear, and buffalo-grass, while in the Rocky Mountains it frequents dizzy 

 heights among the rocks just below the timber-line. It occurs along with cacti in 

 abundance all over the southern portion of the Great Basin, and on into Mexico. 

 All of this region is extremely barren. 



Discopleura cajrillacea, DC. — While not limited strictly to sand-hills, this species 

 seems to be at least characteristic of sandy districts. Have collected it on sandy 

 wastes, in Cloud county, and in various places in Sedgwick county. 



Viola tricolor, L.; var. arvensis, DC. — This species is extremely abundant all 

 over the more sandy portions of Sedgwick county. Of course it is not at all limited 

 to sandy regions, but seems only to be well adapted to such localities, and to some 

 extent characteristic of them. It is often associated with Primus chicasa, and is 

 found on the outskirts of low woodlands and along roadsides and rising ground. 



Cristatella Jamesii, Torr. — I have collected this species at Concordia, Cloud 

 county, on sandy knolls near the Republican river, and have found it in abundance 

 in the vicinity of Wichita, also, but always in sandy soils. 



Linaria Canadensis, Spreng. — Have collected very few specimens of this species ; 

 always in sandy wastes. According to Gray's Manual, it is common in sandy soils- 



—3 



