MONOSOMICS 317 



its first gametophytic mitosis. Sub-haploid pollen-grain divisions 

 have, however, been found in Uvularia (Belling, 1925 c) and in an 

 Allium hybrid (Levan, 1936). 



These observations show that while embryo-sacs and animal 

 gametes, lacking a part of the haploid complement, may function 

 (on account of their very limited activities) the male gametophyte 

 of plants is usually incapable of development in this condition. 

 Further, the zygote of plant or animal can never develop when 

 lacking either a section or a whole chromosome of one type entirely, 

 i.e., in both sets. The organism is slower and poorer in develop- 

 ment where a section or a small chromosome is lacking in one set, 

 and non-viable where a larger element is lacking, e.g., one of the 

 large autosomes in Drosophila. 



The statement is satisfactory for general purposes, but it requires 

 two qualifications, one of which is obvious from what has already 

 been said. The haploid complement probably always contains 

 reduplications. We might then expect that it would not all be 

 necessary or equally necessary to the life of the organism. That 

 this is so has been proved by Muller (1935). Certain deficiencies of 

 small segments are not lethal in Drosophila in the homozygous 

 condition. The segments lost are presumably present elsewhere. 

 The organism is protected against their loss as a polyploid is. The 

 second quaUfication depends on some parts of the chromosome being 

 inert ; its consequences will be discussed later. 



The fact that a certain group of materials is necessary for the 

 successful working of the life processes enables us conveniently to 

 define the haploid complement of an organism as the minimum 

 combination of chromosomes or parts of chromosomes necessary in 

 single or double number for the development of the organism to 

 maturity. 



(b) Progeny of Triploids. The second kind of evidence is to 

 be derived from the progeny of triploids. In these the extra 

 chromosomes are distributed at random at the first and second 

 divisions (with slight loss through lagging) shown by Mather (1935) 

 with trivalents in Triticum, and therefore gametes are produced 

 with an approximately binomial frequency distribution of chromo- 

 somes ranging from the haploid to the diploid number (Table 47). 



