438 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1912. 



which had become lost — a resuming of the abandoned lines of develop- 

 ment of the forefathers. It was on account of this that these rever- 

 sions or atavistic characters could be carried to a high or at least 

 sufficiently balanced degree. If these had been newly acquired char- 

 acters and transmissible in pure strains, their diversion in the Mende- 

 lian sense would nevertheless have been impossible, for I have been 

 convinced of this recently by completed control experiments with 

 animals which had enjoyed the peculiar adaptation for a short period 

 only. In these, neither of the two peculiarities of the parents is com- 

 pletely dominant in the offsprings, but they stand midway between 

 their producers, and the acquired character pales, fades, becomes 

 enfeebled, and is finally lost. 



I do not wish to close my observations without considering another 

 phase of the inheritance of acquired characters, namely, the trans- 

 mission of acquired disease and the transmission of acquired resistance 

 against toxin. 



Strictly considered in the light of experimental investigation, we 

 possess only the classical experiments of Brown-Sequard, Westphal, 

 and Obersteiner that deal with transmission of disease. These were 

 responsible for epilepsy produced in the guinea pig by operation. 

 Brown-Sequard and Obersteiner severed some cords of the spinal 

 cord, or more frequently the hip nerve in their guinea pigs; Westphal 

 tapped them on the head with a hammer once or several times. 



The most marked results of these interferences consist in epileptic 

 cramps, as well as an occasional change of the eyeball, namely, a 

 whitish dulling of the cornea and protrusion. These changes pro- 

 duced by disease are found again in a part of the progeny; even the 

 second generation possesses at times a tendency to epileptic attacks 

 from the first without a repetition of the operation. These experi- 

 ments have since been twice tested; first, by Sommer with completely 

 negative results; but this can scarcely be considered conclusive on 

 account of the limited amount of material experimented upon; and 

 second, by Macisza and Wrzosek, in which the older observations 

 were completely verified and their scope only slightly restricted by 

 the fact that incomplete attacks could be produced in a few healthy 

 young, from normal parents, a fact not observed by any of the previous 

 investigators. 



As regards the inheritable transmission of protection against bac- 

 terial or other toxins, we have the pioneer experiments of Elirlich, 

 now famous for his remedy for syphilis, on mice protected against 

 rizin and abrin; also those of Tizzoni and Cattaneo on mice protected 

 against tetanus and rabbits protected against hydrophobia; finally, 

 those of Behring upon rabbits protected against diphtheria; but these 

 are not quite beyond challenge, since the transmission of toxin resist- 

 ance upon the progeny was then observed only when both parents, or 



