NOVEMBEB 15, 1907] 



SCIENCE 



659 



Of course, the association of sex-distinc- 

 tion with an accessory chromosome is ad- 

 mittedly a peculiarity of certain types, but 

 science proceeds by the discovery of pre- 

 rogative instances, of which surely this 

 notable illustration will long be remem- 

 bered. 



While knowledge has of late progressed 

 so rapidly in regard to many genetic phe- 

 nomena, we still know next to nothing of 

 the facts relating to the incidence of 

 partial sterility among heterozygous forms. 

 Guyer found that the abnormality of 

 which the sterility of hybrid pigeons is the 

 expression, begins in the reduction-division 

 and is apparent as an entanglement of the 

 chromosomes which fail to divide. In 

 many eases sterility is partial; and for 

 example, a proportion of good pollen- 

 grains occurs mixed with the aborted 

 grains. Puller examination of these cases 

 would probably lead to interesting results. 



In selecting facts which tend to limit 

 our outlook on the phenomena of heredity 

 I have naturally chosen to speak rather of 

 features which are positive and mutually 

 consistent than of the many negative and 

 thus far conflicting items of evidence 

 which must perhaps one day be allowed 

 their weight. The real value of these 

 negative and frequent doubtful observa- 

 tions is as yet so uncertain that they must 

 be regarded rather as hints to be followed 

 in the pursuit of facts than as facts al- 

 ready ascertained. 



Allelomorphism, as we are becoming 

 more and more disposed to believe, con- 

 sists in the separation of a positive some- 

 thing from the absence of that something: 

 More correctly, perhaps, we should say that 

 the thing which conveys a certain power 

 segregates, leaving in that cell division no 



types followed peculiar but consistent systems, are 

 evidently to be referred to the eftects of coupling, 

 as Doncaster has shown. 



representative of that power behind. This 

 allelomorphism is the one fact of which we 

 have the clearest proof. It may govern, 

 as we have seen, features of the utmost 

 diversity. What then is that allelomor- 

 phism? An essential phenomenon of cell 

 division, it is not: for in homozygous 

 organisms the products of division are 

 alike. Any theory of heredity must in- 

 clude and recognize both these two kinds 

 of division in its purview. We seek vainly 

 as yet for a scheme by which these two 

 sorts of division may be represented. 



I do not know that analogy is helpful 

 in these cases, but in my own mind I some- 

 times remember in this connection that the 

 somatic divisions themselves are also of 

 two types. There are segmentations 

 which, as in radial animals or bilateral 

 animals, divide similar parts from each 

 other, and there are also the serial divi- 

 sions by which series of differentiated 

 segments are produced. It seems to me 

 just possible that the heterogeneity among 

 the diiTerentiated segments may have some 

 point of real resemblance to the hetero- 

 geneity of allelomorphs. I suggest this 

 comparison with only a faint hope that 

 it may prove sound. 



Lastly, any scheme of heredity must be 

 able to recognize the possibility of gametic 

 coupling between allelomorphs belonging 

 to distinct pairs, and though few such 

 couplings have yet been proved, we have 

 good reason to believe that yet other 

 systems of couplings of much higher com- 

 plexity exist. 



Dr. Loeb encourages us to look to chem- 

 istry for the fulfilment of our hopes, and 

 often, as in the case of the sweet peas, of 

 which I have spoken, we come very near 

 indeed to something like simple chemical 

 . phenomena. Of chemistry I know little, 

 but I would ask those who are experts in 

 chemistry whether it is in harmony with 



