94 



SCIENCE 



[N. S. Vol. XXXII. No. 811 



then to expect to increase this proportion only 

 if he can multiply artificially the effectiveness 

 of nature's causes; and it seems hardly reason- 

 able to be disappointed if positive results are 

 not obtained from experiments with only 1,200 

 or 1,500 plants. 



One method which in spirit is an extension 

 of the injection work was suggested by Os- 

 borne's investigations on plant proteids. Work 

 on the ultimate composition of pure proteids 

 has only been touched, but the fundamental 

 researches that Osborne and his associates 

 have carried on for the past twenty years have 

 shown, even with the crude methods of our 

 general analytic chemistry, that the proteids 

 of different species of plants are very different 

 in composition, the differences becoming more 

 definite as the plants are further apart in the 

 natural system. These facts immediately sug- 

 gest the possibility that if the plant of one 

 species could in its first life stages utilize the 

 stored proteids of the endosperm or cotyledons 

 of a very different tj^pe, changes would prob- 

 ably be induced in it, some of which might 

 be heritable. This treatment is quite different 

 from that where plants are fed different quan- 

 tities of inorganic compounds in the form of 

 the so-called essential elements of soil fertility. 

 Food compounds and enzymes or producers of 

 enzymes of a different kind from those ordi- 

 narily produced and used by the plant, are 

 ready for its use in the very early formative 

 period of ontogeny. If any changes can be 

 expected to come about indirectly through 

 changes in nutrition, they should be expected 

 to occur under such treatment. It is possibly 

 not a phenomenon that could occur naturally, 

 yet since variations are caused by some cell 

 activity different from the normal, they might 

 very well be caused by the production of a 

 different proteid or part of a proteid molecule, 

 different from that normally produced but 

 similar to what is produced by other plants. 

 Then again similar conditions are probably 

 produced when severe changes in temperature 

 occur during the maturation of the seed. In 

 fact, abnormal temperature conditions seem 

 to have similar effects on somatic cells, for 

 Webber'^ once stated that after the last great 



^ In a personal communication. 



Florida " freeze " bud variations were very 

 numerous in the trees that were severely in- 

 jured. It is likely, too, that specialized para- 

 sites may have had changes in their own struc- 

 ture brought about in this way. Of course 

 one must recognize the fact that a great many 

 data are at hand on the effect of the stock 

 upon an engrafted scion, all of which are 

 negative or questionable. In this case, how- 

 ever, conditions are not similar. The bud or 

 branch used as a scion is not at all in the 

 same ontological stage as are the young seed- 

 lings in the experiments proposed. I have 

 mentioned this matter at greater length than 

 I had intended, but I have had the chance to 

 try only some preliminary experiments with 

 grafted cotyledons, and hope the plans might 

 find favor with others who could work on a 

 larger scale. 



The second experiment which included ob- 

 servations on 1,200 plants is somewhat differ- 

 ent. There seemed a possibility at least that 

 if certain characters are due to the presence of 

 an enzyme, this enzyme might actually be 

 added to a plant which had not inherited 

 either the enzyme itscK or the ability to pro- 

 duce it, if the proper time of addition and the 

 proper method could be found. This is pure 

 speculation with no analogies, although it is 

 likely that the mosaic disease of tobacco is an 

 enzyme effect that can be communicated by 

 infection. This disease, however, is not in- 

 herited through the seed, and may prove to be 

 bacterial. 



The subjects selected for experiment were 

 two varieties of tomatoes, Sutton's Best of All, 

 a variety with red flesh, and Golden Queen, 

 a variety with yellow flesh. Hurst had al- 

 ready shovm that all tomatoes possessed yellow 

 flesh and that presence and absence of red 

 flesh acts as a simple Mendelian pair. In 

 order to be c'ertain that we are dealing with 

 the same characters, however, the two varieties 

 were crossed, and Hurst's work corroborated. 

 Golden Queen, the yellow variety, was grown 

 to flowering in sterilized soil. Flowers were 

 castrated and bagged. When the stigmas were 

 receptive they were pollinated with pollen from 

 the same plants, which had been kept pure by 

 bagging. At intervals of one hour after pol- 



