372 
MiscelkDiea 
probability that tluU children remain longer at ischool than very intelligent ones, and start later. 
These factors act in a not easily intei'pretable way on the number of pairs of dull and very dull 
children. 
On the other hand Dr Gordon's pk)tted observations show little more than the variations due 
to i-andom sampling in ,sueh a slender series. Both sets of observations together undoubtedly 
indicate within the limits of error one and the same law of relationship*. It is almost impossible 
to conceive that such diverse environmental conditions rather than a fundamental germinal 
relation could produce such concordance. The conclusion Avhich is emphasised hy material 
drawn l)y such different methods from such very difterent environments is that the relation of 
intelligence lietween siblings is fixed b^' something more innate than environment. That some- 
thing more innate, more constant and more universal in its domination can only be the hereditary 
factor. 
Of course the results in the jiresent paper for the relation between Intelligence Quotient and 
Mentace can only be considered as suggestions until we have far longer series of pairs of siblings 
tested by the Binet-Simon or allied methods. But they serve to indicate that very fruitful work 
can be achieved in this direction, and even the present data owing to their relatively limited 
environmental conditions niay lielp io disjiel the n(_>tion — largely based on prejudice, not on 
acquaintance with actual measurements — that differential environment is the source of resem- 
blance between siblings. 
II. Variation and Distribution of Leaves in Sassafras. 
By N. M. GRIER. 
The following note is made on the basis of e.\aminati(jn of ten sassafras trees and 102 seedlings 
near Pittsburg, Pa., and eight trees near St Louis, Mo. Only three kinds of leaves were met 
with, three-lobed, two-lobed and single-lobed, but it may be infei'red that the same laws will 
govern the distril)ution of the four- five- and six-lobed forms described by Berry some years ago in 
the Botanical Gazette. 
The single-lobed leaves are in great prepondei'ance, constituting two-thirds of the foliage in 
Pittsburg specimens, while, near St Louis, three trees were observed in which other than single- 
lobed leaves were wanting. In these an extensive self-pruning had taken place. The terminal 
lea\'es of young branches are single-lobed, although there may be an occasional two-lobed leaf. 
Tops of trees are usually comijosed almo.st entirely of single-lobed leaves. 
The dissected forms of lea\'es aj^pear to be most plentifully developed under the influence of 
sViade. In such cases they were most thickly distributed at the middle of the tree (as has been 
noted for three-lobed leaves in the Britton and Bi'own Flora), on young twigs whose terminal 
leaves were dissected, and toward the bottom on older twigs. There was a tendency for more 
three-lobed and less one-lobed lea\'es to be found on smaller twigs growing near the trunk, but 
occasionally on larger twigs, or smaller boughs gi-owing among the larger boughs. 
No transitional forms between the three-lobed and two-lobed leaves were noted on the same 
tree. The latter apparently increase in lunnber as the three-lobed forms decrease, and are 
associated mostly with the single-lobed leaves, being about equally distributed between the 
younger and older twigs. They are rarely found ac the top of the tree. Evidence that the 
available amount of light may play some part in the distribution of leaves is found in the fact 
that the great majority of observed seedlings growing in the shade dex'elop the three-lobed or 
two-lobed leaves in combination. Contrast is offered by a statement made in a standard 
American textbook of Ijotany — " In Sassafras, almost any leaf may be entire or variously lobed, 
* Both series of observations also indicated how satisfactorily the normal law of distribution may 
be applied to material of this kind. 
