738 



SCIENCE 



[N. S. Vol. XXXII. No. 830 



and came true to type, but then died out. 

 No hybridization experiments were made 

 with them, but in other similar experi- 

 ments, which I have not time to mention in 

 detail, modified forms produced by the 

 action of changed conditions gave typical 

 Mendelian characters when crossed with 

 unmodified decemlineatas, thus proving 

 that the induced characters were constant 

 and heritable according to the regular laws. 



I have thought it worth while to relate 

 these experiments at some length, because 

 they seem to me to be very important, and 

 because they do not appear to have at- 

 tracted the attention in this country that 

 they deserve. 



They are confirmed to a very large ex- 

 tent by the experiments of Professor Klebs 

 on plants, the results of which were pub- 

 lished this summer in the Croonian Lecture 

 on "Alterations of the Development and 

 Forms of Plants as a Result of Environ- 

 ment." As I have only a short abstract 

 of the Croonian Lecture to refer to, I can 

 not say much on this subject for fear of 

 misrepresenting the author; but, as far as 

 I can judge, his results are quite consistent 

 with those of Tower. Sempervivum 

 funckii and 8. acuminatum were subjected 

 to altered conditions of light and nutrition, 

 with the result that striking variations, 

 such as the transformation of sepals into 

 petals, of petals into stamens, of stamens 

 into petals and into carpels, were produced. 

 Experiments were made on Sempervivum 

 acuminatum, with the view of answering 

 the question whether such alterations of 

 flowers can be transmitted. The answer 

 was in the affirmative. The seeds of flow- 

 ers artificially altered and self-fei'tilized 

 gave rise to twenty-one seedlings, among 

 which four showed surprising deviations of 

 floral structure. In two of these seedlings 

 all the flowers were greatly altered, and 

 presented some of the modifications of the 



mother plant, especially the transformation 

 of stamens into petals. These experiments 

 are still in progress, and it would perhaps 

 be premature to lay too much stress upon 

 them if it were not for the fact that they 

 are so completely confirmatory of the re- 

 sults obtained by similar methods in the 

 animal kingdom. 



I submit to you that evidence is forth- 

 coming that external conditions may give 

 rise to inheritable alterations of structure. 

 Not, however, as was once supposed, by 

 producing specific changes in the parental 

 soma, which changes were reflected, so to 

 speak, upon the germ-cells. The new evi- 

 dence confirms the distinctions drawn by 

 Weismann between somatic and germinal 

 variations. It shows that the former are 

 not inherited, while the latter are; but it 

 indicates that the germ may be caused to 

 vary by the action of external conditions in 

 such a manner as to produce specific 

 changes in the progeny resulting from it. 

 It is no more possible at the present time to 

 connect rationally the action of external 

 conditions on the germ-cells with the 

 specific results produced in the progeny 

 than it is possible to connect cause with 

 effect in the experiments of Herbst and 

 Stockard ; but, when we compare these two 

 kinds of experiments, we are no longer able 

 to argue that it is inconceivable that such 

 and such conditions acting on the germ- 

 plasm can produce such and such effects 

 in the next generation of adults. We must 

 accept the evidence that things which ap- 

 peared inconceivable do in fact happen, 

 and in accepting this we remove a great 

 obstacle from the path of our inquiries, 

 and gain a distinct step in our attempts to 

 discover the laws which determine the pro- 

 duction of organic form and structure. 



But such experiments as those which I 

 have mentioned only deal with one aspect 

 of the problem. They tell us about ex- 



