122 bulletin: museum of comparative zoology. 



the offspring all possible combinations of the three kinds of chromo- 

 somes. Or, if mating involving any two of the types could be made, 

 there should result all possible combinations between them. On 

 account of the small number of animals available for my study, no 

 conclusions as to whether these conditions are realized in nature could 

 be drawn. The presence of the three types in these few animals, 

 however, strongly suggests the possibility of realization, especially 

 since two of the three possible combinations are realized for the two 

 kinds of chromosome in type Ci. The presence of a third type also 

 suggests that there may exist in this case the mechanism for the 

 transmission of triple allelomorphs. 



b. Some experimental Evidence. 



The most extensive breeding experiments the results of which tend 

 to show that the chromosomes are concerned in the transmission of 

 hereditary characters are those on Drosophila by Prof. T. H. Morgan 

 and his students. In the course of this work they have dealt with 

 over a hundred unit-characters which show Mendelian inheritance, 

 either in a typical or modified form. In Drosophila, there are four 

 pairs of chromosomes, of which one pair is very small, and one is a 

 pair of heterochromosomes, or " sex-chromosomes." In their behavior 

 in inheritance, the hundred and more characters fall into four groups, 

 each group tending to behave as a unit, just as it would be expected 

 to do in case it were carried by a single pair of chromosomes. Of 

 these groups of characters, one is very small, the others much larger, 

 the largest one being the group of " sex-linked " characters. Naturally 

 the small group of characters has been correlated with the small pair 

 of chromosomes and the group of sex-linked characters with the sex- 

 chromosomes. 



But there have been exceptions in the case of many pairs of allelo- 

 morphs, especially those that are sex-linked, i. e., cases where factors 

 belonging to a certain group have gone into a mating together, but 

 have not always reappeared together, as they would be expected to do 

 if they were all carried by a single chromosome and that cliromosome 

 maintained its individuality. These phenomena have been explained 

 by the so-called "cross-over" hypothesis. In this connection Morgan 

 ('11) developed what has been termed the "linear arrangement" 

 hypothesis, which was further elaborated by Sturtevant ('13). These 

 authors assume that the factors, or "genes," which represent the 



