458 THE AMERICAN NATURALIST [Vol. XLIV 



The stains that we employ do little more than differen- 

 tiate basic from acid bodies and in this regard the chro- 

 mosomes belong to the acid group. Their finer differ- 

 ences, if such exist, would not appear by the methods 

 used. 



The most striking evidence that can be cited to show 

 that the chromosomes are different is based on their size 

 relations. These are constant. Does this mean that the 

 chromosomes are therefore different? I do not believe 

 that such evidence is of any value one way or the other. 

 If the size of the chromosomes is referable to their 

 genetic continuity, the facts can be accounted for without 

 recourse to the assumption of chemical difference. 



Fortunately we have some evidence from embryology 

 that has seemed to many embryologists to indicate that 

 the chromosomes differ in their physiological behavior; 

 from which we may infer that they differ chemically. 

 I refer to Boveri's brilliant experiments with the disper- 

 mic eggs of the sea urchin. 



When two spermatozoa enter simultaneously the egg 

 of the sea urchin each brings in its own center or aster 

 from which two centers are formed. These two centers 

 form a triaster (one being excluded) or a tetraster about 

 the three pronuclei (two male, one female). When the 

 nuclei dissolve each sets free its 18 chromosomes, pro- 

 ducing 18 X 3 = 54 chromosomes which are distributed 

 to the three poles of the triaster, or to the four poles of 

 the tetraster. The distribution is, as a rule, irregular 

 in the sense that some centers get more than others. 

 The protoplasm then divides into three or into four equal 

 parts, the axis of division corropoiidinn' with that of the 

 egg axis as in normal division. 



From these eggs embryos develop; many of them are 

 abnormal, but a few are normal. Normal embryos de- 

 velop more often from the eggs that divided at once into 

 three, than from those that divided into four. Boveri 

 points out that the chance is greater in the three-fold 

 type that each cell gets at least one set of the chromo- 



