470 



The Journal of Heredity 



and may not he a case parallel to those 

 described in sunflowers.^ 



A c:iia>(;ei) <;ladi()ias 



Diagram of the flower wilh segments re- 

 versed, described in footnote to the text. 

 This reversal of orientation did not affect 

 the shape of the flower nor a change in the 

 relative location of the dark marking. It 

 is evident, therefore, that these things are 

 due to independent factors. (Fig. 12.) 



It is natural to recall the brilliant 

 researches of Bridges, showing that 



irregularities in chromosome distribution 

 produce the most far-reaching results. 

 Thus in Drosophila the individual with 

 one X-chromosome is a male, with two 

 a female, but that with three is not 

 viable. In the fertilization of plants 

 it seems that a quantitative difference 

 decides what shall be embryo and what 

 endosperm. Can such differences arise 

 in somatic mitosis? My colleague, Dr. 

 R. C. Whitman, assures me that the 

 cancer irregularities in mitosis, resulting 

 in inequalities in the distribution of 

 chromatin, have actually been observed.^ 

 I take this opportunity to state that 

 the title of my paper in the Journal 

 OF Heredity, August, 1917, should have 

 been simply "Sunflower Seedlings." The 

 first three words, which are not appli- 

 cable, were added after the proof left 

 mv hands. 



Decline of Germany's Birth Rate During the War 



In German cities, it is stated, the 

 birth rate for 1916 was nearly 40% 

 lower than that for 1914. Infant 

 mortality was higher than us^^ual. J-im- 

 ilar conditions proljably ])revail in 

 most belligerent countries, although in 

 England the infant mortality decreased 

 during the first two years of the 



war. It seems likely that the infant 

 mortality ^^ould be highest among the 

 poor and ignorant, but it is conceivable 

 that the decline of the birth rate may 

 be greatest among the most highly 

 educated and carefullv nurtured women. 

 If so, it will be a serious dysgenic 

 effect of the war. 



' Another case, evidently belonging to a different category, is that of a pink-flowered gladiolus 

 now before me. The lateral inferior segments of the perianth are smaller than the others and 

 are strongly bicolored, the base deep pvirplish-rose, the ajMcal part (except the pale pink margin) 

 clear pale yellow. On one spike, while the upper flowers are normal, as just described, the lower 

 ones have only one of these modified segments, which is central. It overlaps the two large 

 lower ones, which are hke the upper; whereas in the normal flower the large median lower segment 

 is overlapped by the bicolored ones. In this case there has occurred a reversal of the orientation 

 of the perianth segments, so that instead of having two outer above and one below, and one inner 

 above and two below, there are two outer below and one above, and two inner above and one 

 below. This has not afTected the general shape of the flower, nor has it caused a reversal of the 

 orientation of the dark marking, affording a curious illustration of tlie independence of the three 

 factors. 



*I have recently found a garden Canna, with flowers normally bright yellow, the petals pro- 

 fusely spotted with bright red, in wliich one flower on a spike was al>normally colored. In this 

 flower two of the petals were entirely brilliant red, while a third had exactly one-half red, and 

 the other half yellow with the elongate red spots. The dividing Tne is exactly in the middle, 

 and is absolute. This is quite similar to the bicolored sunflower ray, and it is especially inter- 

 esting as showing in a single jjetal the two types of bicoloration: one (the spotting) due to a pat- 

 tern factor, normal to the plant; the other one to cell mutation or somatic segregation. On 

 the segregation hyixjthesis we should ajjparently be compelled to supjx)se that the full red tyjje 

 was recessive to the spotted. Docs any reader know whetlier this is the case in Cannas? 



It is also to be remarked that the .somatic mutation or segregation evidently took place at 

 an early stage in flower formation, and l)oth the canna and the simflowcr indicate that tlie foun- 

 dation of any single petaloid element is biccllular. The canna petal must have lieen derived 

 from contributions from two already different types of cells, one destined to produce the red, 

 the other the s()ottin^. The e.'cactly e jial contributions in li.^ate in early bicellular condition, 

 not the fusion of two multicellular groups. 



