ON VARIATIONS. 363 



of the nervous system in guinea-pigs. As Brown- 

 Sequard experimented over a period of thirty years on 

 thousands of guinea-pigs, it might be thought that we 

 could accept his results as absolutely conclusive. Yet 

 a repetition of some of his experiments by Romanes 

 and by Hill seems to show that they may be very 

 largely erroneous. Thus, like Brown-Sequard, Ro- 

 manes * found that some of the progeny of parents in 

 which an injury to the restiform body had produced 

 protrusion of the eyeball, showed a protrusion likewise, 

 though this was less marked, and always affected both 

 eyes; but it seemed that this might be an accidental 

 occurrence, in that normal guinea-pigs are sometimes 

 to a certain extent exophthalmic. Again, Romanes 

 found that some of the progeny of animals in which 

 hsematoma and dry gangrene of the ears had super- 

 vened after injuring the restiform body, also became 

 affected. However, the morbid state seemed to arise 

 at any time in the life history of the individual, and the 

 process not only affected a much less quantity of the 

 ear, but also a different part of it. One therefore 

 might imagine it to be due to mere coincidence, or to 

 transmitted microbes; but Romanes does not think this 

 can be the case, as, on the one hand, he has never seen 

 the peculiar morbid process of the ears in other guinea- 

 pigs, and, on the other hand, he was unable to inoculate 

 the ears of healthy animals with matter from the ears 

 of mutilated guinea-pigs. 



Romanes repeated Brown-Sequard's experiments on 

 the section of the cervical sympathetic nerve, but he 

 never observed in their progeny any change in the shape 

 * " Darwin and after Darwin," vol. ii. p. 104 et seq. 



