332 Proceedings of Royal Society of Edinburgh. [july 4 , 
of birtli of the females of the family, and places the names of 
the sisters in a family after the names of all the brothers, but it 
also systematically omits all mention of children who died young ; 
and when children have grown up to maturity and died unmarried, 
they also are often omitted. 
A careful examination of various Peerages has left upon my 
mind the general impression that too much reliance should not 
be placed upon individual facts contained in them. There are 
many sources of error, which are in practice not sufficiently guarded 
against. Sometimes an obvious misprint is made in the edition 
for one year, and is repeated without correction in the editions 
for several successive years. In a few cases there seems good 
ground for believing Foster’s statement, that the information 
furnisht by members of the peerage, has been intentionally 
incorrect. Occasionally, though very rarely, the marriage of a peer 
or his son is mentioned in one Peerage and not in another ; and the 
same is the case with regard to the issue of some marriages. These 
inaccuracies, however, are not sufficiently numerous to produce any 
appreciable effect upon the general results of an enquiry of the 
present kind ; and I think that, if proper precautions are adopted, 
the vital statistics furnisht by the records of the British Peerage, 
are more complete and trustworthy than we can hope to get in 
almost any other way. Perhaps better statistics might be got 
from the records of some of the Widows’ Funds which grant 
benefits to the children, as well as to the widows, of members ; but, 
in the absence of these, I think the Peerage statistics the best we 
have. 
Taking now a general survey of the facts supplied by the Peerage, 
we see that they relate to different classes of persons, and that all 
the facts so supplied are not equally trustworthy. We may, I 
think, assume that, in general, the facts relating to each peer will be 
the most complete and trustworthy ; that those relating to his 
children during his life will be almost (or quite) as trustworthy; and 
that the information will become less trustworthy in proportion as 
the relationship to the peer of the day, is more distant. We have, 
in the case of every peer included in our list, the necessary informa- 
tion as to his sons, and their sons (if any), also as to his father, 
his brothers, and their sons (who are the peer’s nephews). In 
