DIVISION OF CHROMOSOMES 47 



across the mitotic stage, out of the shades of one resting nucleus 

 and into the shades of another equally unknowable. They have a 

 continuity which depends on the permanence of the linear arrange- 

 ment of their characteristic constituents seen as a loosely coiled 

 thread at early prophase of mitosis and as a closety coiled spiral 

 thread at metaphase. This conclusion is of the first importance in 

 genetics. The old dictum, oninis cellula e cellula, may be extended 

 to nuclei, to chromosomes and to the unit-particles making up the 

 chromosome, cytologically known as chromomeres, genetically as 

 genes. It is even truer to sa}^ that all genes arise from pre-existing 

 genes than that all cells arise from pre-existing cells, in this sense 

 that the gene is an older structure than the cell. 



The permanent thread is double at mitotic prophase. In the 

 early prophase meiosis, as we shall see, it is single. We must there- 

 fore assume provisionally that it is always single at telophase. 

 The one change that it undergoes during the resting stage is that of 

 longitudinal division, presumably following the reproduction by 

 each permanent particle or chromomere of one like itself, and the 

 absence of this division distinguishes the resting stage before 

 meiosis {cf. Chs. IV and X). 



It seems that the centromere, on the other hand, does not divide 

 at the same time as the thread itself, and has not the property, 

 which the chromatids have, of associating in pairs. The special 

 properties of the centromere will be clearer from a study of meiosis 

 and of the results of irradiation (Chs. IV and XII). 



4. ABNORMAL FORMS OF MITOSIS 

 Deviations from the described type of mitotic division are found 

 in the higher plants and animals, but they are nowhere such as to 

 render the process unrecognisable. In the lower organisms marked 

 deviations from the described type occur (Fig. 10). These are 

 described by Belar in a comprehensive survey of mitosis in the 

 Protista (1926 5). 



In many Fungi and Protista, an intranuclear spindle is found. 

 In this type the centrosome, which is characterised by the presence 

 of permanent granules colourable with nuclear stains, may lie 

 outside the nucleus, as in some AmcebcB, or inside it, as in the 



