No. 640] VARIATION AND HEREDITY IN LUPINUS 439 



pipersniithii, three pink apricus, and three dark blues. 

 The seeds since used have all been derived from these 

 nine plants, each of which has been assigned a pedigree 

 number. 



In 1918 seeds of each of the 1917 plants were again 

 planted. Although about 400 seeds of each were planted 

 in each culture, two pedigrees failed to yield a single plant 

 that year, though a few plants came up the following 

 spring. Pedigree VIII, Light Blue, yielded a culture of 

 85 plants and pedigree V, Striped White, 28 plants. Thus 

 the highest per cent, of germination did not exceed about 

 20 per cent., ranging from that down to zero. 



In 1919 a series of germination tests were carried out. 

 This had not been ])Ossible before on account of the small 

 number of seeds available. A considerable variety of 

 methods wore employed. Soaking the seeds in tap and 

 distilled water for weeks is of little value, owing to 

 the failure of the seed to imbibe water. Breaking or 

 cutting the seed coat brings about prompt imbibition of 

 water and consequent swelling, but does not produce a 

 higli ]ior('ontage of germination. Seeds were soaked in 

 watci- iindfi- air pressures up to 140 pounds to the square 

 inch with no noticeable effect. 



Since the seeds are small cutting or filing the seed 

 coats is a very arduous affair where cultures of any size 

 are to be grown. Attempts were, therefore, made to find 

 some other means of bringing about the same result. The 

 seed coat appears to be diflicnlt to wet and this wn< 

 thought to be due to the prescnct^ «>l' some oily or waxy 

 constituent. Attempts to dissolve tliis by means of KOII, 

 various ]iercentages of alcohol, ether, etc., i)roved en- 

 tirely nnsnceessful. 



Soaking in ("on('(»ntiated ILSO^ proved the most effica- 

 cious method tried. This was applied in parallel series 

 of seeds from the same plant for i)eriods from 5 minutes 

 up to 2 hours. The shorter treatments seemed to produce 

 no effect at all. The longer i>eriods of two hours or 

 over killed the enil^ryos. After some experimentation it 



