100 



SCIENCE 



[N. S. Vol. XXXIII. No. 838 



anism of environic responses has had any- 

 thing more than the most limited useful- 

 ness. The stimuli of climatic and many 

 other agencies do not imply the introduc- 

 tion of any strange or new substances into 

 the bodies of the organs affected. These 

 agencies might change the dissociations in 

 such manner as to modify the relative num- 

 ber of free ions and thus alter the molec- 

 ular complex of the living matter in a very 

 important manner however. The intricate 

 play of enzymatic action might also be 

 altered, and any modification of the rela- 

 tive reaction velocities of the more impor- 

 tant processes might result in material and 

 permanent change, especially in those 

 cases, in which external agencies interfere 

 directly with the action of the germ-plasm. 



The introduction of solutions into ovaries 

 or the exposure of reproductive elements 

 to unusual irradiation may raise the addi- 

 tional liability of disturbed polarity and of 

 modified surface tensions in the cells. It 

 is conceivable that the rearrangement or 

 disturbance of the localizations of sub- 

 stances, especially the mineral salts, might 

 seriously modify the capacities of the bear- 

 ers of heredity. These direct and material 

 possibilities offer an adequate basis for the 

 organization of experimental research upon 

 the main subject as well as the means of 

 interpretation of results without recourse 

 to schemes of particulate inheritance or 

 theories as to the constitution of germ- 

 plasm to which may be ascribed usefulness 

 in the discussion of other problems in evo- 

 lution. 



The theoretical considerations which 

 might lead us to assign all cases of per- 

 petuation of environic effects to the direct 

 action of the exciting agency upon the 

 germ-plasm are perfectly competent to be 

 questioned with each new bit of evidence 

 brought forward. My own results so far 

 have been all of this kind, as are many of 



the instances cited, but the case does not 

 appear so simple when, as in Sempervi- 

 vum, somatic alterations are induced late 

 in the ontogeny and are transmitted by 

 seeds borne on the altered branches, and 

 equally serious doubts are raised when one 

 considers the multifold somatic alterations 

 of Capsella and the fact that they must be 

 repeated dozens, perhaps hundreds, of 

 times before being transmissible. One 

 must not lose sight of the fact that the 

 soma is itself the most closely interlocking 

 environment of the germ-plasm, and not 

 until the germ-plasm has been exposed to 

 interaction with a changed soma for re- 

 peated generations does it undergo the 

 changes necessary to an altered heredity. 



The problems of real interest, however, 

 are those which concern the actual infiu- 

 ence of environment upon evolutionary de- 

 velopment, and with regard to these many 

 generalizations of importance are becoming 

 possible, and their brief summarization 

 may suggest some encouraging advances. 



First it is to be seen that not all en- 

 vironic effects induced in the laboratory or 

 by transplantations are heritable, although 

 they may be carried over for two or three 

 generations. No satisfactory basis has yet 

 been found upon which it might be pre- 

 dicted that any effect would be ephemeral 

 or permanent. The characters induced in 

 an hereditary line may be regressive, or 

 awakened latencies or organizations de 

 novo. The alterations which become per- 

 manent may be cumulative in construction, 

 although they are mutative or abrupt in 

 their appearance in most instances. Ab- 

 ruptly displayed by some organisms, yet 

 they may not become heritable until the 

 germ-plasm has been repeatedly subjected 

 to the direct action of the exciting agency 

 or to the effect of a changed soma for many 

 generations. 



Orthogenetic or heterogenetic as to direc- 



