IV.] FOUNDATION OF A THEORY OF HEREDITY. 191 



that, in his argument, Strasburgcr has over-estimated the 

 support afforded by exact observations upon indirect nuclear 

 division. Certainly the fact, discovered by Plemming, and 

 more exactly studied by Balbiani and Pfitzner, that, in nuclear 

 division, the loops split longitudinally, is of great and even of 

 fundamental importance. Furthermore, the observations, con- 

 ducted last year by van Beneden, on the process of fertilization 

 in Ascaris, have given to Flemming's discovery a clearer and 

 more definite meaning than could have been at first ascribed 

 to it. The discovery proves, in the first place, that the nucleus 

 always divides into two parts of equal quantity, and further 

 that in every nuclear division, each daughter-nucleus receives 

 the same amount of nuclear substance from the father as from 

 the mother ; but, as it seems to me, it is very far from proving 

 that the quality of the parent nucleoplasms must always be 

 equal in the daughter-nuclei. It is true that the fact seems to 

 prove this ; and if we remember the description of the most 

 favourable instance which has been hitherto discovered, viz. 

 the process of fertilization in the Q.gg of A scans, as represented 

 by van Beneden, the two longitudinal halves of each loop cer- 

 tainly impress the reader as being absolutely identical (com- 

 pare, for instance, loc. cit., Plate XIX, figs, i, 4, 5). But we 

 must not forget that we do not see the molecular structure of 

 the nucleoplasm, but something which we can only look upon 

 (when we remember how complex this molecular structure 

 must be) as a very rough expression of its quantity. Our most 

 powerful and best lenses just enable us to make out the form of 

 separate stainable granules present in a loop which is about to 

 divide : they appear as spheres and immediately after division 

 as hemispheres. But according to Strasburger, these granules, 

 the so-called microsomata, only serve for the nutrition of the 

 nuclear substance proper, which hes between them unstainable, 

 and therefore not distinctly visible. But even if these granules 

 represent the true idioplasm, their division into two exactly 

 equal parts would give us no proof of equality or inequality in 

 their constitution : it would only give us an idea of their quanti- 

 tative relations. We can only obtain proofs as to the quality of 

 the molecular structure of the two halves by their effect on the 

 bodies of the daughter-cells, and we know that these latter arc 

 frequently different in size and quality. 



