EXPLANATORY HYPOTHESES 121 



transmitted, but that after they have been trans- 

 mitted several times they become constant, is 

 strongly in favour of Professor Bering's hypothesis ; 

 for this improvement by repetition is characteristic 

 of a mental operation. Physical forces act with as 

 much certainty the first time as afterwards ; repeti- 

 tion makes no difference to them. But in order 

 that instincts may work regularly, it is necessary that 

 they be constantly repeated ; that the stimulus 

 which starts the movements should not be absent 

 too long. If the stimulus does not come at the 

 right time the cell cannot respond ; and if the 

 stimulus remains in abeyance for a very long period , 

 the cell, or group of cells, may forget how to respond 

 even when the stimulus does come. This gives 

 rise to retrogression or degeneration. On the con- 

 trary, the more an organ is stimulated the more 

 vivid will be the memory in the next generation. 

 The organ will be better developed, and this will 

 give rise to definite variation. This appears to me 

 to be the true explanation of the Lamarckian doc- 

 trine, that use or disuse are the cause of change in 

 organs , and the true reason why acquired characters 

 in time became congenital. 



So also, when the germ contains two different 

 memories, derived from its two parents, these may 

 clash and antagonise each other, and so allow an 

 older and dormant memory to be stimulated into 

 activity. This is atavism. Or degraded characters, 

 which have suffered from disuse, can, on a renewal 

 of the old stimulus, again be recalled; as we see in 

 Proteus, which gets dark in colour when kept in the 

 light. 



