434 EXPERIMENT STATION RECORD. [Vol.43 



A third duplication of generic factors in shepherd's purse, G. H. Shuix 

 (Ahs. in Science, n. ser., 51 (1920), No. 1328, p. 506). — In continuation of tlie 

 author's study on generic factors of shepherd's purse (E. S. 11., 41, p. 134), a 

 report is given of a third generation of a cross between a wild biotype of the 

 common shepherd's purse (Bursa bursa-pastoris) from Wales and the alli(>d 

 species B. heegeri. In this cross there appeared a small number of plants of 

 unique type, having a more coriaceous texture than either of the plants in- 

 volved in the cross. This type has been designated as coriacea, and it appeared 

 in the Fa family in almost exactly the ratio of 1 : 15. 



Subsequent breeding experiments have shown that this type breeds true 

 •when selfed and have also confirmed the interpretation of this as a third case 

 of duplication of factors in this species. 



FIELD CROPS. H 



[Work with field crops at the North Carolina Station in 1918-19], 



C. B. Williams (North Carolina Sta. Rpt. 1919, pp. 32-39).— The different 

 lines of work with tobacco conducted at the Granville farm are briefly noted. 

 The results of general fertilizer tests are reported as indicating that acid phos- 

 phate is the best source of phosphoric acid for flue-cured tobacco, as bone meal 

 and basic slag retard maturity. The use of 36 to 40 lbs. of potash per acre 

 even at the very high cost of the material showed a profit, but larger quanti- 

 ties were not so profitable. Good tobacco was grown also with the use of 

 barnyard manure and hardwood ashes, as well as with tobacco stems and phos- 

 phoric acid. It was found that by closer planting with intensive fertilization 

 the yield could be doubled practically and the quality maintained. Experi- 

 ments in harvesting by priming the leaves and cutting the stalks have been 

 noted heretofore (E. S. R., 38, p. 37). Six yeai's' work has shown definitely 

 the advantage of priming over cutting. It is stated that tobacco seed beds can 

 be maintained indefinitely provided the beds are properly sterilized each year. 



In cotton-breeding experiments during the past two years the highest yield- 

 ing strain on the average produced 146 per cent more seed cotton than was 

 secured from the lowest yielding strain. Other similar average variations 

 reported are 59 per cent more seed cotton per boll, 33i per cent greater length 

 of lint, and 17 per cent greater percentage of lint to seed. When cotton plants 

 from Mississippi and North Carolina grown seed from the same pure strain 

 were grown in comparison, the plants from the ]\Iississippi seed showed slightly 

 heavier yields, a little longer staple, and somewhat earlier maturity. 



In connection with crop improvement work the following varieties and strains 

 are regarded as valuable for distribution : Biggs Seven-ear corn ; Haberlandt 

 No. 38, Virginia No. 11, and Mammoth Yellow No. 101 soy beans ; King No. 29 

 and Mexican Big Boll cotton ; Leap Prolific No. 12 wheat ; and Abruzzi rye. 

 The breeding of soy beans for high oil content resulted in strains producing 

 higher percentages of oil, but thus far these have been the poorest yielders, 

 and the largest quantity of oil per acre has been found to come from those 

 strains which yield the largest quantity of seed. 



Cooperative experiments with the U. S. Department of Agriculture in seeding 

 wheat and oats at different rates and on different dates are briefly described. 

 In all plantings with wheat up to December 1 the 60-lb. seedings gave the 

 best yields. The highest yields of grain at the Iredell farm were secured 

 from seeding oats at the rate of 30 lbs. per acre on October 1 and November 1, 

 of 60 lbs. on October 15, and 120 lbs. on November 15. The best yield of straw 

 was secured from the seeding made November 1. At the Edgecomb farm the 



