E. Schuster 
479 
greyhounds '507 {Biometriha, Vol. il. p. 379, Table V.). The general conclusion is 
expressed as follows : — " The coefficient of parental heredity varies to some extent 
from character to character in man, having a mean value of about "46. This value 
is however in quite close agreement with tbe results obtained for other species, 
and we may roughly say that parental heredity in the species hitherto dealt with 
is close to 'o." {Biometrika, Vol. ii. p. 396 (v.).) 
The value obtained for fraternal correlation is less satisfactory, but if it is not 
directly confirmatory of, it is in no way inconsistent with previous determinations 
for other characters, for the range of variation of fraternal correlation seems to be 
much greater than that of parental. Thus "74 here arrived at is much larger 
than the mean value for other characters in man ("5), but it does not differ much 
from the values obtained for coat-colour in horses and dogs, or for certain characters 
in waterfleas and aphides, all of which will be found duly set forth in the paper 
referred to above. 
One more conclusion can be drawn, and that is the urgent necessity for the 
collection of satisfactory data, on which to work at this and other problems in 
heredity. The Interdepartmental Committee on Physical Deterioration, in their 
report, published in 1904, strongly recommend that an Anthropometric Survey 
should be undertaken by the Government at the earliest possible moment (vide 
Section 54). Should this recommendation ultimately be adopted, there is no 
reason why the scope of this survey should not be somewhat extended so as to 
include the collection of data bearing on heredity ; but till then one must make 
what use one can of any material that one can get. 
Postscript. Since writing the above I have been able to obtain values for 
paternal, maternal, and fraternal correlation from information collected in England. 
For the first two the following material was used : 
(1) A record of the offspring of marriages in which one or both parents were 
deaf from the register of the National Association in Aid of the Deaf and Dumb. 
Of 161 marriages there recorded both parents were deaf in 127, while in 26 the 
husband was deaf and the wife was either hearing or not stated to be deaf, and in 
8 the reverse was the case. Thus there were in all 153 marriages in which the 
husband was deaf, of which 125 had proved fertile by the time of the information. 
The total number of children was 435, of which 52 occurring in 25 families were 
deaf. There were 135 marriages in which the wife was deaf 109 of these had 
resulted in a total number of 379 children, of which 50 occurring in 25 families 
were deaf The only part of this used in the tables was that in families with a 
deaf father there were 383 hearing children and 52 deaf, and in families with a 
deaf mother there were 329 hearing children and 50 deaf 
(2) The rest of the material used is contained in Table XVIII. The number 
of deaf parents given there is probably much understated. Both Dr Kerr, to whose 
kindness I owe the information concerning the L.C.C. schools, and Mr Moore, who 
sent me that relating to Margate, warned me of this. 
