xx INTRODUCTORY LETTER TO SENATOR MOSSO 



different species of organisms. The weighings were 

 made in the morning before the animals were fed. 

 But they were kept always supplied with dry oats ; 

 this practice is desirable because it helps essentially 

 in preserving the animals in good condition. It does 

 not entail a sufficient error in the weights to be ob- 

 jectionable, because it is more or less constant and is 

 not very large, as the animals will not eat a great 

 deal of grain when they have plenty of other food. 

 No fresh food was left in the pens or boxes over 

 night. 



o 



In all the weighings there is necessarily an error. 

 A positive error, because the digestive tract, particu- 

 larly the wide ccecum, contains always considerable 

 quantities of undigested matter ; moreover, the blad- 

 der may hold a greater or less quantity of urine. 

 A negative error, because every illness, even a very 

 slight indisposition, and every injury, such as a bite, 

 for instance, causes a greater or less loss of weight. 

 The quantitative values of these errors are presum- 

 ably not very great ; they probably counterbalance 

 one another to a certain extent in the averages, which 

 may be accepted as approximately accurate. 



The advantage of these experiments over statistics 

 taken from man lies especially in the fact that the 

 same individuals are followed through the whole 

 period of growth. Otherwise we may reach errone- 

 ous conclusions ; thus in girls there is a very great 

 acceleration of growth during the two or three years 

 preceding puberty, that is, the epoch of the first men- 



