BEHAVIOR OF SPECIES OF CACTI. 19 



general nature. This is probably accounted for, in most cases at 

 least, by the cold nights of April and May. It should be mentioned 

 that the season of blossoming is a month earlier at Brownsville than 

 at Chico. In general, all species are inclined to be a little more gray 

 in the fall and the young growth a little more highly colored in the 

 spring at Chico than at either of the other stations. The purplish 

 tinge so common in species of the flat- jointed group is as unstable a 

 character as it is in other groups of plants, in spite of the fact that in 

 some descriptive works it is considered of fixed value. Purplish tints 

 may be induced at will in many species. The subjection of the plants 

 to low temperatures will often cause them to turn brilliant in color, 

 while many assume the same aspect when the root system is injured, 

 or from other causes if they suffer from drought. 



PROLIFERATION OF FRUIT. 



There are a number of species in this genus which have under nor- 

 mal conditions proliferous fruits, that is, new fruits rise from the old 

 ones, which become incorporated as a permanent part of the plant, 

 thus making pendent bunches of fruits of greater or less extent. 

 Sometimes the fruit remains attached indefinitely but is not prolifer- 

 ous, as in the common white-spined cane cactus of the highland of 

 Mexico ; 1 or it may be proliferous and form large pendent bunches, 

 as in the cholla of southern Arizona, 2 Opuntia prolifera of southern 

 and Lower California, and the South American species 3 so exten- 

 sively introduced into Australia. One cane cactus of southern Ari- 

 zona 4 is not proliferous fruited in its native heath, but when grown 

 under cultivation in a more humid atmosphere in California it is very 

 commonly proliferous fruited, and this character is more especially 

 brought out in that form of the species from the borderland between 

 the typical form and its closest relative, the Colorado-New Mexico 

 cane cactus. 5 It is not at all uncommon to find bunches of a dozen 

 fruits of this form all attached in some of the California collections. 

 The truly desert form does not appear to assume this habit as readily. 

 At Chico and San Antonio only the desert forms have been grown, 

 and in neither situation have they yet taken on this habit. 



In some instances there appears to be some correlation between 

 this proliferation and sterility. The fruit of the southern Cali- 

 fornia cholla 6 is usually sterile. Sometimes a single seed is pro- 

 duced which no one, so far as known, has ever caused to germi- 

 nate. The fruits of the Arizona cholla 7 apparently have normal em- 

 bryos and are abundantly produced, but no one has as yet reported 



1 Opuntia imbricata. 5 Opuntia arborescens. 



2 Opuntia fulgida and O. mamillata. e Opuntia prolifera. 



3 Opuntia monacantha. 7 Opuntia fulgida. 

 * Opuntia spinosior. 



