January, ]922. 



SCIENTIFIC AGRICCLTrKf:. 



157 



Variation and Inheritance in Red CI 



M. 0. MALTE, Ph.D. 

 Doiuiiiion Agrologist, Central E xporiinoiital Farm, Ottawa. 



over 



III.— CORRELATIOX OF 

 CHARACTERS. 



{Concluded from Last Two Issues.) 



Numerous observations have been made 

 which have led to the assumption that cn-- 

 tain characters may be correlated, i.e., 

 hereditarily bound together and therefore 

 appearing associated with each other. 

 Time of Development and Color of 

 Flowers. 



Holdeflciss (1) observed a retarded de- 

 velopment in plants that were white- 

 flowered, the white-flowered plants blos- 

 soming as much as four weeks later than 

 the red ones (Exp. Sta. Rec. Vol. 31, 1914, 

 p. 830). W^iiether this observation, how- 

 ever, can be used in support of the con- 

 tention that a correlation may exist be- 

 tween time of development and color of 

 the flowers, is very doubtful. Comment- 

 ing on the subject, Kajanus (5, p. 3) is in- 

 clined to regard the coincidence of late 

 flowering and white flowers in HoldefJeiss' 

 plants as accidental, as he thimself has 

 never observed any striking difference in 

 blossoming time between white and red 

 flowering forms. 



Color of Seed and Color of Flowers. 



As white-flowering red clover plants 

 generally have yellow seed free from 

 purple, it has been asserted that white 

 flower colour and yellow seed colour may 

 be correlated. Thus Holdefleiss (1) main- 

 tains tjiat such correlation exists. Gernert 

 (1 p. 88) observes that it was found that 

 white-flowered plants always produced 

 yellow (or pea) seed; he adds, though, 

 that some of the pink and light-red flow- 

 ered plants also produced yellow seed, while 

 the seeds from dark red flowers were in- 

 variably dark purple. Finally Gmelin (2, 

 p. 466) also maintains that w^hite flower 

 color probal)ly always is coupled with yel- 

 low seed color. 



This is, however, not the case. In the 

 first place ordinary red-flowering plants 

 may produce all kinds of seed, varying 

 from yellow to deep purple. And, se- 



condly, w/iite-flowering plants have Ijcen 

 observed which have purely white seed. 

 Wliite-seeded, white-flowering plants have 

 thus been found by Dr. H. Witte at the 

 Plant Breeding Station of Svalof, Sweden. 

 Such a plant was also found by Dr. M-ittr 

 and the writer in 1919 at a small railway 

 station in S. Dakota. 



Concerning other correlations between 

 seed color and flower color a very interest- 

 ing observation made by Kajanus (3 p. 767, 

 and 5 , p. 3) may be mentioned. .Kajanus 

 has found that blue flower color seems 

 always to be associated with a peculiar 

 orange-brown seed color. He remarks that 

 all blue floM^ered red clover plants w/iic/i 

 so far (up to 1914) have been iiarvested 

 have had orange-brown seed. 



Color and Weight of Seed. 



A great number of observations bearing 

 on the relation between weight and color 

 of red clover seed have been made Ijy 

 various investigators. Preyer (1, p. 13^1 

 found when examining three commercial 

 samples, that the purple seed in all the 

 cases was heavier than the yellow. Fru- 

 wirth (2) came to a similar conclusion and 

 so did Kajanus (3, p. 769) and Eastman 

 (1, p. 98). Haherlandt (1) and Von 

 BumTier (1, p. 204) however are on record 

 as ha\ang found yellow seed picked out 

 from commercial samples to be heavier 

 than purple. 



In the above cases the results were se- 

 cured from mixed bulk samples. Suet 

 results, however, are, as will be readily 

 understood from the. discussion in the 

 following, rather unsatisfactory as indi- 

 cators of the existence of any real genetic 

 correlation between color and weight. . 



Investigating the weight of differently 

 colored seed from individual plants 

 Bauman (1, p. 66) found tjiat the weight 

 of the seed of yellow seeded plants in so- 

 called Bucheggberg Matten clover was 

 lower than the weight of the seed harvested 

 from dark-seeded plants. Holdefleiss (1) 

 also asserts that in the material investi- 



