47 



o\ar\- l)Ul sometinic's oiu- of the- carpels is onh' ])aiiiall\- (k'\i'l- 

 opcd, rcsultinti in a somewhat irregular sti^iua and oiiK' two 

 complete cells. In one case a flower with two separate pistils 

 was observed. The pistil, however, shows the least variation in 

 si7-e of all the organs of the llower of E. propullans, the style being 

 uniformh- 5 mm. in length and the ovary about 3 mm. when the 

 flowers are in anthesis. 



So far as the writer is aware the flowers of /;;. propullans are 

 considerably smaller than in any other species of the genus. 

 They \'ary in length from 9 to 13 millimeters but the majority 

 are a little over 10 mm. long. Attempts to ascribe to other 

 factors than heredity the difference in the size of flowers of var- 

 ious related species of a genus are mostly futile, yet the hypothe- 

 sis put forth by Blodgett* in explanation of the reduction in the 

 size of the flowers of E. propullans seems to the present writer 

 at least very plausible. 



As is well known the offshoot in E. propullans pushes out from 

 the stem near the middle (Fig. 3), its bud originating "at the 

 base of the peduncle in the axil of one of the leaves." Thus the 

 "vascular system of the peduncle supplies, through branches, 

 the necessary strands for the offshoot." This side-tracking of a 

 considerable amount of the food supply going up the peduncle 

 may have had, in the opinion of Blodgett, "considerable influ- 

 ence in the reduction in size noticeable in the flowers of this form 

 in contrast to the rest of the genus." 



It seems very probable that the prevalent reduction in the 

 number of the floral organs is due to the same cause and we have, 

 at least in this species, a very simple physiological explanation 

 for the fluctuations in the floral structures. 



In conclusion it is worth noting that the genus Erythronium 

 belongs to a subfamily of the Liliaceae in which the trimerous 

 plan of the flower is quite consistently adhered to. The char- 

 acteristic variations in the number of the perianth segments, 

 stamens, and carpels and especially the preponderance of dimer- 

 ous flowers in E. propullans are therefore very striking. 



* Blodgett, F. H. The Stem Offshoot in Erylhronium propullans Gray. Johns 

 Hopkins Universitj' Circular, 3-5. June, 1909. 



