On the Dimegalous Sperm and Chromosomal Variation ot Euscliistus, etc. 141 
most strangely of all he enters as proof of chromosomal variability the 
different coimts reached by different observers on the same object, whereas 
such differences are almost certainly due to differences in the personal 
eqnation and degree of observational experience. In the latter manner 
he treats the cases of Artemia and Anasa. 
On the whole, therefore, we may fairly say that many of the cases 
from the literature adduced by Della Valle in favor of his views should 
be stricken from the list for the reasons just given. At the same time 
there are certain cases that cannot be tlms explained away, and we will 
return to them later. 
Della Valle presents a detailed criticism of the methods of coim- 
ting chromosomes, and concludes by preferring the smear method to 
sections, and the study of cells in thin transparent membranes to cither. 
This criticism w^ell deserves reading, for accurate countiiig of chromosomes 
is most difficult as all those understand who have had the experience. 
He maintains that the coimts of those who Support the theory of the in- 
dividuality of the chromosomes are scarcely to be trusted on account of 
theii’ methods, yet he does not hesitate to eite from the literature in Support 
of his views many observations based upon such evidcnce. I would re- 
})eat again that the method of using thick sections is in my experience 
preferable to the smear method, and that those coimts probably have 
the most accuracy wdiich deal wüth small numbers of chromosomes for the 
difficulty in coimting increases in geonietrical ratio wüth the increase in 
the number of the chromosomes. 
Della Valle then enumerates and criticizes what he calls „sub- 
sidiary hypotheses for making the aberrant cases conform with the theory 
of the individuality of the chromosomes”. By these he means the expla- 
nations of those who maintain there is a normal number, and who try to 
account for variability on other grounds than ,, chromosomal lability”. 
Thus when a Variation in number is found higher than the normal, some 
students, as 1 in my present account of EuscJiisius, have argued that it is 
due to the failure of conjugation of certain univalent chromosomes. Della 
Valle criticizes this by saying that the number of chromosomes in the 
first maturation is taken as the basis of argument; but he is surely mistaken 
in this, for as the basis in generally regarded the number in the fertilized 
egg or the oögonia or spermatogonia before bivalent chromosomes have 
beeil produced. ln other cases this has been accounted for by the preco- 
cious division of a iiarticular chrornosome. Della Valle objects that 
if this be due to an equational division it has not been demonstrated, and 
that such an explanation would be admissible only in case every cluomo- 
