April 7, 1905.] 



SCIENCE. 



527 



probably holds good. The three principal 

 kinds of protoplasm of the aseidian egg 

 before cleavage are the ectoplasm which 

 gives rise to ectoderm, the mesoplasm which 

 produces mesoderm and the endoplasm 

 which becomes endoderm. Each of these 

 three substances is derived in part from 

 the nucleus of the egg; the ectoplasm 

 comes from the egg nucleus at the begin- 

 ning of the first maturation division, the 

 other two from the nucleus at an earlier 

 stage in the ovogenesis. In every cycle of 

 division a large amount of chromatic ma- 

 terial escapes from the nucleus into the 

 cytoplasm and I have found in mollusks 

 and ascidians that this substance is then 

 differentially distributed to different areas 

 of the egg and that it gives rise in part to 

 the principal formative substances of the 

 embryo. These facts lend support to the 

 hypothesis of intracellular pangenesis pro- 

 posed by de Vries ; they show that even 

 though a certain number of general dififer- 

 entiations may be transmitted through the 

 cytoplasm, such as polarity, symmetry and 

 localization, nevertheless the mechanism 

 exists for the nuclear control of the cell and 

 they thus afford a means for harmonizing 

 the facts of cytoplasmic organization with 

 the nuclear inheritance theory. 



We find, therefore, that the germ is by 

 no means simple, even if we consider only 

 the visible structures of the cell, and that 

 its organization is sufficiently complex to 

 exercise a determining influence upon de- 

 velopment and evolution. Similarities in 

 the character and localization of the ma- 

 terial substances of the egg must be the 

 initial causes of all similarities or homol- 

 ogies which appear in the course of de- 

 velopment. Modifications of this germinal 

 organization, however produced, are prob- 

 ably the immediate causes of evolution. 



II. If we inqiiire how such modifications 

 of the germ arise and what the particular 

 modification is which is associated with a 



certain mutation of the adult organism, we 

 pass from the region of observed fact to 

 one of hypothesis, for in only a few in- 

 stances have such germinal mutations been 

 observed. Nevertheless, enough is known 

 regarding the organization of the germ cells 

 to warrant our hazarding a 'shrewd guess' 

 as to the nature of these germinal muta- 

 tions. The nuclear inheritance theory 

 points to some modification in the struc- 

 ture, number or distribution of the chro- 

 mosomes or of the elements of which the 

 chromosomes are composed as the initial 

 cause of mutation. The fact that the Men- 

 delian ratio in alternate inheritance cor- 

 responds to the ratio of chromosomal dis- 

 tribution in the maturation and fertiliza- 

 tion of the egg indicates that in such chance 

 distribution of these chromosomes we have 

 the principal cause of the law of alternate 

 inheritance, as Wilson, Sutton, Cannon and 

 others have pointed out. It may also be 

 reasonably inferred that in this chance dis- 

 tribution of chromosomes we have one of 

 the most potent causes of individual varia- 

 tions. 



It is well known that the disappearance 

 of characters does not necessarily imply 

 their final loss ; many heritable qualities 

 remain latent through one or more genera- 

 tions, only to appear as active characters in 

 subsequent generations. In such cases it 

 is probable that the material bearers of 

 these qualities also remain latent, though 

 we are wholly ignorant of what constitutes 

 latency as contrasted with activity in chro- 

 mosomes. 



Relatively little is known as to the fac- 

 tors which determine the number of chro- 

 mosomes or as to the effect of varying num- 

 bers on adult organization. Montgomery 

 concludes from his studies on the Hemip- 

 tera that certain chromosomes are in the 

 process of degeneration and disappearance 

 in these animals. In this case the heritable 

 qualities which were borne by these chro- 



