458 Prof. J. B. Farmer. On the [May 13, 



composed of individuals derived from the male and female parents respec- 

 tively could hardly receive stronger support. 



The second example I will take from a hybrid fern which is at present 

 under investigation in my laboratory. The plant in question, Polypodium 

 Schneideri, is the result of a cross between P. aureum and P. vulgare var. 

 elegantissimum. The nuclei of the former contain 35 chromosomes after meiosis, 

 whilst P. elegantissimum has about 97 or 98. The chromosomes of the hybrid, 

 at meiosis, instead of exhibiting a mean between the two parents, i.e. 66, has 

 a much larger number. There is a little variation in the actual numbers, 

 but the average lies between 90 and 100. That is to say, the quantity 

 characteristic of the elegantissimum parent is retained. Besides this, we find 

 at meiosis a large number of small, unpaired, chromosomes ; these obviously 

 represent the surplus over and above those furnished by the aureum parent. 

 At division they betray their real nature by passing in their entirety to one 

 of the daughter-nuclei, when these are formed. 



Facts such as those I have recounted seem to emphasise the permanent 

 individuality of the chromosomes, and indeed they have often been appealed 

 to in this sense. The case is further strengthened by the not unfrequent 

 occurrence in cells, before they have undergone meiosis, of one or more pairs 

 of chromosomes that are different from the others, a circumstance to which I 

 have already alluded. 



Moreover, the mode of sorting out, and the general behaviour of the 

 chromosomes throughout the cell-generations of the organism, are at first 

 sight in striking accord with the results of statistical investigations on 

 heredity, and they are looked upon by many people as representing the 

 primordia of the characters which we have already considered. 



But I believe there is a fatal objection to this point of view, and one which 

 conclusively proves that we cannot regard the chromosome itself as the 

 structural unit responsible for the characters of the organism. The objection 

 is based on their relatively small numbers, and it was urged, though in 

 a somewhat different connection, by Weismann, many years ago. I will 

 illustrate my meaning by one example, and I wish to express my thanks to 

 my friend, Mr. A. D. Darbishire, for kindly collecting the facts of the case. 

 In the two pure races of Pisum sativum and P. arvense, there are no less 

 than 18 pairs of characters respectively which in the hybrids behave as 

 allelomorphs. Since, however, there are only seven chromosomes in the 

 nuclei of the gametes, it is evident that the 18 allelomorphs could not all 

 occur independently of each other, but that, if the chromosome be adopted 

 as the unit, the allelomorphs would appear in at most seven groups, and 

 further, that the members of each group would always appear in association. 



