o 



10 PROCEEDINGS OF THE AMERICAN ACADEMY 



tains : fl. May to September. — Stems 2-6 feet high, at the base 

 1-3 inches thick ; joints- 3-3^ lines in diameter, ashy gray ; the 

 singular flattened and angular tubercles 2^-3 lines long ; spines H- 

 2 inches long, crowded together at the upper end of each year's 

 growth, very loosely sheathed. Flower purple, half an inch in diam- 

 eter. Fruit 9-10 lines long, covered with reddish-brown bristles. 

 Seed 2 lines or less in diameter. 



* 



* 



* The material for the present study of our Cactacece is not as 

 full as would have been desirable in the examination of so difficult a 

 family. Hence it may sometimes have happened, that what I have 

 endeavored to distinguish as species are forms which properly belong 

 together ; or I may have combined as one species incomplete speci- 

 mens of quite distinct plants. The fear of confusing heterogeneous 

 plants under one name, and the desire to indicate to future explorers 

 all the different forms known to me, combined to induce me to pro- 

 ceed as I have done. 



For those who naturally may be horrified at the idea of 117 species 

 of Cactacece. in a territory where, a few years ago, scarcely half a 

 dozen were known, I will indicate how the mass of material may be 

 comprehended under fewer types. 



Of Ma7)iillarice the species 1-9 are quite distinct, and can in no 

 manner be united ; 10-12 might perhaps be considered as forms of 

 a single species; 13- 17 are all very distinct; 18 and 19, 20-23, 25 

 and 26, 27 and 28, may possibly be forms of only four types, instead 

 of 10, as I have enumerated them, thus referring my 30 species to 

 22 types. 



In the genus Echinocactus the following species might be united : 

 1 and 2, 7 and 8, 9 and 10, 12 and 13, 14 and 15, — leaving 15 in- 

 stead of 20 types. 



The following species of Cereus will perhaps bear reduction : Nos. 

 1 and 2, 3 and 4, 5-7, 10 and 11, 12-14, 16 and 17, 18-22 

 (though some of them, of which I do not even know the flowers, may 

 prove to belong even to different sections !), 23 and 24, — thus redu- 

 cing my 31 species to 18 types. 



Ojmntia is a still more difficult genus, and mistakes are here most 

 easily made. Many of them are as yet very incompletely known ; 

 and without being able to compare a great number of living specimens 



