';-() Lutz, Triploip Mutants in (knothera. 



6. Unreduced g failing to complete the second division -(- 

 unreduced cT. 



7. Unreduced failing to complete the second division -|- 

 unreduced cf failing to complete the second division. 



It will be apparent to the reader from the fore-going that there 

 is little left to worry about unless he is in search of a number 

 that will disprove the theory that somatic chromosome numbers 

 owe their origin to the behavior of the germ cells. 



It is not necessary to attempt an estimate of the relative 

 frequency with which these various combinations should be expected 

 to occur. It would be an impossible undertaking. But it will be 

 at once apparent that by far the greater number should come under 

 the head of I. 2; after that a very much smaller number should 

 appear under I. 4 and III. 2, and perhaps a still smaller number 

 under V. 2 or III. 4. Further than that I will not attempt to guess. 

 Many, of course, are far beyond the limits of probability. 



While combinations with germ cells failing to complete the 

 second division have been included in this list, it is hoped that it 

 will be recalled that very little emphasis is laid upon these. They 

 are mentioned merely because of the fact that it is possible to 

 explain certain somatic chromosome numbers on the basis of the 

 assumption that the germ cells sometimes fail to complete the 

 homotypic division. 



If, in addition to the above, we consider the possibility of one 

 or more chromosomes occasionally failing to be included within 

 the daughter nucleus or daughter nuclei at the telophase of the 

 heterotypic or homotypic mitosis, the list of possibilities for somatic 

 chromosome numbers could be much extended. A further com- 

 plication would be added should we later find that chromosome 

 degeneration may occasionally occur in any form having a number 

 of chromosomes in excess of 14. Evidence will be brought out in 

 the next paper to show that there is some indication of this in 

 both lata and yiyas. 



Gates, in July, 1907 (8, p. 13), seeking an explanation for 

 the chromosome conditions which he had reported for what was 

 then supposed to be offspring of O. lata X 0. Lamarckiana (lata, 14; 

 u Lamarckiana hybrid", 20 or 21) offers among a number of others 

 the following suggestions: 



"(2) It is possible that 0. lata might produce two kinds of eggs, 

 having respectively seven and fourteen chromosomes. If both these 

 kinds of eggs were fertilized with (). Lamarckiana pollen and produced 

 embryos, we should have plants resulting with fourteen and twenty- 

 one chromosomes. The difficulty here, however, is that on such 

 an hypothesis the union of seven lata chromosomes would produce 

 a lata plant; while the union of fourteen lata chromosomes with 



