'268 PEESIDENTTAL ADDRESS. 



tions, and whether there is any indication of such high 

 action being inherited. By differential weighting, however, 

 a differential action might be established, and any hereditary 

 tendency would be the more readity observable. Prelimin- 

 ary experiments in this direction on rapidly breeding ani- 

 mals would probably soon lead to indications of the most 

 convenient form of continuing the experiments and observa- 

 tions. In single-lop rabbits there is a decided twist or set 

 to the skull. There is no distinct evidence, at present, 

 whether this is due to correlated variation or to the in- 

 herited effects of the lop. It would be possible, I suppose, 

 to induce a lop in other strains of rabbits, and perhaps in 

 guinea-pigs, and that quite painlessly by artificiallj^ weight- 

 ing the left ear. A number of the rabbits or guinea-pigs 

 •might be at the outset divided into two groups, which should 

 be kept as far as possible under similar, conditions, except 

 that in one set the left ear of each individual was kept 

 slightly weighted. After four or live generations, certain 

 individuals of the weighted group should be set aside, 

 unweighted from their birth, and observed ; while observa- 

 tions and measurements of the skulls of some of them should 

 be made and recorded, and carefully compared with those 

 made on the skulls of individuals weighted from birth. If 

 it should be found that at the end of twenty or thirty 

 generations, though the effects on the individual were well 

 marked, there was no evidence of any inherited effect, such 

 negative evidence would be pro tanto of value, while positive 

 evidence obtained in this way does not seem to be open to 

 damaging criticism. 



Satisfactory proof of the inherited ejBfects of increased 

 use by itself will, I imagine, be difficult to. obtain under 

 experimental con li tions ; observations on the effects of 

 disuse, by itself, wiuld be complicatsd by the panmixia 



