CuAP. VII. SUMMARY, SALTS OF AMMONIA. 173 



why we should reject it as incredible. Prof. Donders, 

 of Utrecht, informs me that from experiments formerly 

 made by him and Dr. De Knyter, he inferred that less 

 than the one-millionth of a grain of sulphate of atro- 

 pine, in an extremely diluted state, if applied directly 

 to the iris of a dog, paralyses the muscles of this organ. 

 But, in fact, every time that we perceive an odour, we 

 have evidence that infinitely smaller particles act on 

 our nerves. When a dog stands a quarter of a mile to 

 leeward of a deer or other animal, and perceives its 

 presence, the odorous particles produce some change in 

 the olfactory nerves ; yet these particles must be in- 

 finitely smaller * than those of the phosphate of am- 

 monia weighing the one-twenty-millionth of a grain. 

 These nerves then transmit some influence to the brain 

 of the dog, which leads to action on its part. With Dro- 

 sera, the really marvellous fact is, that a plant without 

 any specialised nervous system should be affected by 

 such minute particles; but we have no grounds for 

 assuming that other tissues could not be rendered as 

 exquisitely susceptible to imj)ressions from without if 

 this, were beneficial to the organism, as is the nervoiLs 

 system of the higher animals. 



* My son, George Darwin, has from g^fro to Trrowij of an inch 



calculated for me the diameter of — in diameter. Therefore, an ob- 



a sphere of phosphate of ammonia ject between -^ and ^ of the 



(specific gravity 1-G78), weigh- size of a sphere of the phos- 



ing the one-twenty-millionth of phate of ammonia of the above 



a grain, and finds it to be -j^ of weight can be seen under a high 



an inch. Now, Dr. Klein informs power ; and no one supposes 



me that the smallest Micrococci, that odorous particles, such as 



which are distinctly discernible those emitted from the deer in 



under a power of 800 diameters, the above illustration, could be 

 are estimated to be from • 0002 to 



•0005 of a millimetre — that is, croscope 



