AMERICAN MEN OF SCIENCE 



803 



only one child and consequently more 

 first-born children. But it appears from 

 the table that the first-born child is more 

 likely than the othei*s to become a scien- 

 tific man. In families of two or more 

 children, 284 are first-born and only 168 

 second-born ; in families of three or more, 

 214 are first-born and 114 are third-born in 

 families of four or more, 159 are first-born 

 and 81 fourth-born. 



Table XIV. The Order of Birth of Scientific 



Men 



The first-born child has been reported to 

 be more likely to be a man of genius, an 

 idiot, a consumptive and various other 

 things, but usually as the result of some 

 statistical fallacy. If, for example, in our^ 

 figures only families of two and eight are^ 

 considered, there being on the average five \ 

 children and one first-^born, it looks at 

 first sight as though the chances are one 

 in five that an individual selected at ran- 

 dom would be the first-born. But both 

 families being represented, it has been 

 argued that the chances are half that the 

 individual is the first-born in the small 

 family and one eighth in the large family, 

 or on the average five sixteenths, nearly 

 one in three. However, the family of 

 eight is four times as likely to be repre- 

 sented in a random selection from the two 

 families, so the chances of the individual 

 being first-born are reduced to one in five. 

 Finally and this seems to have been 



overlooked there are three times as many 

 families of two as of eight, and there are 

 in fact 16 chances out of 61, or a little 

 over one in four, that the individual 

 selec'pd at random will be the first-bom. 



These errors have been eliminated by 

 the method used in the table, but there 

 are it least two other statistical errors 

 that affect such data. "When we are deal- 

 ing vith contemporary' families the older 

 child-en may be preferred. Thus chil- 

 dren under fifteen are not likely to suffer 

 fron pulmonary tuberculosis or to be 

 crimnals, and the patient or the prisoner 

 may be more likely to be the first -bom 

 chile than the last born. This circum- 

 stance holds for the scientific men. They 

 are >n the average 38 j^ears old when they 

 attan their position, and the scientific 

 popilation is rapidly increasing, so the 

 earler born children are more likely to be 

 inclided. This factor, however, is small, 

 as 3 shown by the fact that the second- 

 bori child is not appreciably more likely 

 to )e a scientific man than those later 

 bori. Probably a large statistical error is 

 intioduced by the fact that a man some- 

 tiiiBS does not know of or ignores children 

 whi were born and died before he himself 

 wa born, and thus records himself as the 

 firi-bom. 



n so far as it may in fact be the case 

 th;t the fii-^t-born child is more likely to 

 bea scientific man, this is probably due to 

 social rather than to physiological causes. 

 Gilton found the eldest son to be pre- 

 fered, and perhaps this might be expected 

 uider a sj^stem of primogeniture. In this 

 ccnitry where families are apt to improve 

 tbir economic condition, the younger sons 

 niiy be more likely to be sent to college 

 tlm the older, but as to this there is no 

 a.'iiilable information. 



In Table XV. the intervals are given 

 btween marriage and the birth of the first 

 clild and between the birtlis of successive 

 cHldren for families of different sizes. 

 \hcn there is only one child it is born on 



