My materials were obtained by sending circulars of inquiry to persons who were either twins themselves or near relations of twins.  The printed questions were in thirteen groups; the last of them asked for the addresses of other twins known to the recipient, who might be likely to respond if I wrote to them.  This happily led to a continually widening circle of correspondence, which I pursued until enough material was accumulated for a general reconnaisance of the subject.

There is a large literature relating to twins in their purely surgical and physiological aspect.  The reader interested in this should consult Die Lehre von den Zwillingen, von L. Kleinwaechter, Prag. 1871.  It is full of references, but it is also unhappily disfigured by a number of numerical misprints, especially in page 26.  I have not found any book that treats of twins from my present point of view.

The reader will easily understand that the word “twins” is a vague expression, which covers two very dissimilar events—­the one corresponding to the progeny of animals that usually bear more than one at a birth, each of the progeny being derived from a separate ovum, while the other event is due to the development of two germinal spots in the same ovum.  In the latter case they are enveloped in the same membrane, and all such twins are found invariably to be of the same sex.  The consequence of this is, that I find a curious discontinuity in my results.  One would have expected that twins would commonly be found to possess a certain average likeness to one another; that a few would greatly exceed that average likeness, and a few would greatly fall short of it.  But this is not at all the case.  Extreme similarity and extreme dissimilarity between twins of the same sex are nearly as common as moderate resemblance.  When the twins are a boy and a girl, they are never closely alike; in fact, their origin is never due to the development of two germinal spots in the same ovum.

I received about eighty returns of cases of close similarity, thirty-five of which entered into many instructive details.  In a few of these not a single point of difference could be specified.  In the remainder, the colour of the hair and eyes were almost always identical; the height, weight, and strength were nearly so.  Nevertheless, I have a few cases of a notable difference in height, weight, and strength, although the resemblance was otherwise very near.  The manner and personal address of the thirty-five pairs of twins are usually described as very similar, but accompanied by a slight difference of expression, familiar to near relatives, though unperceived by strangers.  The intonation of the voice when speaking is commonly the same, but it frequently happens that the twins sing in different keys.  Most singularly, the one point in which similarity is rare is the handwriting.  I cannot account for this, considering how strongly handwriting runs in families, but I am sure of the fact.