SECRETARY'S REPORT. 137 



are observations that look as if they had been made with the 

 proper care, and these he will take as the basis of his computa- 

 tion. Now, you must proceed in the same way, and when you 

 read of satisfactory results obtained by some experiment, you 

 must not shrink from the painful investigation as to whether it 

 was made with proper care, and a due consideration of all the 

 elements which should enter into the computation. Therefore, 

 tell your friends, and tell yourselves, when you are satisfied that 

 they and you have made mistakes, that these previous observa- 

 tions are good for nothing, and go to work. Learn to tell your- 

 selves that what you have done is worth nothing, and then you 

 will be on the road of progress. 



It is difficult, but it is the advantage the scientific man has over 

 the practical man. The training of scientific men consists in 

 nothing else but in learning how to set aside their own doings, 

 to criticize their own observations, so that they shall know what 

 is worth listening to and what not. That is the source, of our 

 strength, that is the foundation of our value in community — 

 that we learn (and that is our special office,) how to criticize 

 whatever we do. Now, I think, from what I see here, that you 

 will learn that very soon, and when you have learned that, you 

 will proceed witli confidence. The first thing to eliminate in 

 this experiment concerning the transmissibility of the qualifi- 

 cations of any animal is the ancestral element. 



How will you do that ? By breeding one or two generations 

 in-and-in, without affinity. Here I state a limitation which is 

 not perhaps understood, and I will explain. In order to have 

 stock on which you can make a sound experiment, you must 

 breed together individuals as closely allied as possible, but which 

 shall have no family ties. There is one important element 

 when you speak of breeding in-and-in. I have never heard the 

 distinction referred to that I now make. Breeding in-and-in 

 may mean, according to the way in v/hich I hear it discussed, 

 breeding brother and sister, or father and mother, as well as 

 breeding together individuals which resemble one another very 

 closely. Now, there is a vast difference between these two 

 modes of breeding in-and-in. Breed Anglo-Saxon with Anglo- 

 Saxon, does not mean that brother and sister should intermarry 

 The breed of Anglo-Saxon is improved by the intermarriage of 

 Anglo-Saxons, but of Anglo-Saxons who have no family ties ; 



18* 



