262 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



might be allowed to suppose that the marriages reported were divided according to this ratio, rather than according 

 to the apjiroximate ratio of eqnalitj' found in the asylum. But we should consider that this surplus of non-con- 

 genital deaf would indicate a class who associate principally with hearing persous, and who would, therefore, be less 

 likely to marry deaf-mutes than others would. I think, therefore, that under the circumstances, we should regard 

 the ratio given by statistics of the institution as the most probable one. Of course the reason for this is strengthened 

 if, as you intimate, a large x>roi)ortion of the statistics may be mutual. Allowing for a probable slight tendency of the 

 two classes congenital and non-congenital to choose each other, I think the most probable conclusion would be this: 



Of the congenital deaf one-half married congenital and one-half non-congenital deaf. 



Of the non-congenital three-sevenths married congenital deaf a.ni\ four-seventlts non-congenital deaf. 



And I consider these results sufficiently probable to form the basis of conclusions in cases where slight changes 

 in the numbers would not change the general result. 



If you wish your tal)le returned please inform me. 



Yours, very truly, 



S. NEWCOMB. 



Washington, D, C, May 26, 1884. 



Dear Mr. Bell: The remarkable agreement between the ratio of congenital and non-congenital cases in the 



census reports, and in the numbers married, affords a strong confirmation of the probable soundness of the conclusion 



I indicated to you. The small discrepancy to which you allude probably arose from the twenty-five " not stated" 



cases. I return you the tables. 



Yours, very truly, 



* S. NEWCOMB. 



