DEPARTMENT OF BOTANICAL RESEARCH. 85 



some cases constantly different from each other m such features as 

 depth of green in the f oHage and predominatingly red or green midribs. 

 These experiments throw an interesting light on the amount of diver- 

 sity which may be found by growing the offspring of different indi- 

 viduals of a species occupying the same locahty. 



Another experiment yielding results of considerable interest is a 

 series of F3 families from Oe. ruhricalyx X biennis and the reciprocal. 

 They are for the most part essentially uniform in foliage and pigmenta- 

 tion, having the red buds of ruhricalyx and fohage intennediate in 

 character but nearer the ruhricalyx parent. The inheritance of the 

 difference in flower-size, between ruhricalyx having large flowers 

 (petals about 40 mm. in length) and hiennis having small flowers 

 (petals about 20 mm. in length), is very striking. There is (1) segrega- 

 tion in flower-size between different individuals. In the majority 

 the length of petals is near 25, 30, or 35 mm. ; but in certain plants they 

 are as small as in biennis or (more rarely) as large as in ruhricalyx. 

 Occasionally they far overstep the size of the smaller parent and plants 

 have been found whose petals were all as small as 10 to 12 mm. in 

 length. In addition to this type of segregation there is often (2) 

 striking segregation of flower-size in the same plant. For example, 

 the average length of petals in 9 flowers blooming on the main stem 

 during a period of 8 days was respectively 30, 20.75, 17.75, 31, 15.25, 

 15, 19.25, 28.5, and 31 mm. Furthermore (3) in some cases there is 

 striking segregation, if such it may be called, in the length of petals of 

 the same flower. In rare cases the longest petals of a flower may be 

 twice the length of the shortest. In one extreme case the lengths were 

 respectively 18, 20, 15, and 12 mm. Usually, however, the differ- 

 ence in length of the petals of a flower is less. When it is large it is 

 frequently accompanied by irregularities in the shape or in the color 

 development of certain petals. It is rather surprising to find somatic 

 segregation occurring on such a large scale. 



One other result which may be mentioned is the Fi hj^drid Oe. 

 heivettii X ruhricalyx. Oe. hewettii is a relative of Oe. hookeri described 

 by Cockerell from Colorado. The parent strain used in the cross had 

 red stems and green buds, while Oe. ruhricalyx has red buds and more 

 green stems. There are many other differences, such as width of leaf 

 and character of pubescence. The Fi hybrid is essentially intermediate 

 in all its characters but two, the red buds of ruhricalyx and the heavy 

 pubescence of heivettii being dominant. 



Relationships and Distrihuiion of the Cadacece, by A^. L. Britton and J. N. Rose. 



The taxonomic investigation of the cactus family (by Dr. J. N. Rose 

 and Dr. N. L. Britton) has proceeded uninterruptedly during the 

 year and much progress has been made. The work has been mostly 

 in the greenhouses, museums, and herbaria at New York and Wash- 

 ington, in the preparation of manuscript and of illustrations for the 



