[46] In the case of the egg the whole of the yolk stored by the “oocyte” (cell-generation immediately preceding the maturation divisions) is handed on to only one of the four resulting cells—an obvious economy. The three yolkless cells are necessarily functionless—abortive ova—and are known as the “polar bodies” (Hertwig). In spermatogenesis the maturation divisions, though bearing the same relation to reduction as in oogenesis (Platner, 1889; O. Hertwig, 1890), give rise to four functional germ-cells. The explanation of sexual differentiation given above, and that of polar body formation given here, render it needless to do more than mention the theories of Mimot (1877), van Beneden (1883) and others, by which “maturation” was regarded as removing the “male” element from the otherwise “hermaphrodite” egg.
[47] Weismann postulated a transverse division of the chromosomes, not a distribution of entire chromosomes; but the result as far as the reduction in the number of hereditary qualities goes is the same. The inability of the mitotic mechanism to effect the transverse division of unsplit chromosomes is pointed out by Boveri (1904).
[48] For an exhaustive account of reduction in Invertebrates see Korschelt and Heider, Entwicklungsgeschichte, Allgem. Teil ii. (Jena, 1903).
[49] Nevertheless the possibility of a pseudomitotic interpretation of maturation in Ascaris also has been maintained by O. Hertwig (1890), p. 277, Carnoy and Boveri (1904).
[50] The partial or even complete reconstruction of the nucleus between the heterotype and homotype division in Vertebrates makes it difficult to determine the identity of the split seen in the anaphase of the heterotype with that reappearing in the prophase of the homotype.
[51] e.g. Moore, 1895 (Scyllium); Flemming, 1897; Carnoy and Lebrun, 1899 (Amphibia); McGregor, 1899; Lenhossek, 1898 (mammals), and many others. So also for plants: Strasburger and Mottier, 1897; Dixon, 1896; Sargant, 1896-1897; Farmer and Moore, 1895; Gregoire, 1899; Guignard, 1899, &c.
[52] H. Henking (1899), T. Montgomery (1898) and F. C. Paulmeir (1899) describe the diverging bivalent halves of the tetrad as being united each by two fibres with the corresponding spindle pole. At the next division, at which the diad is resolved into its constituent univalent chromosomes, the daughter chromosomes are attached to the spindle pole each by only one fibre; the two fibres now passing to opposite poles of the spindle being the same fibres which, in the preceding mitosis, were attached to one and the same pole.
[53] Reference may be here made to Rosenberg’s description (1904) of the heterotype mitosis in Drosera hybrids. In the one parent (D. rotundifolia) the somatic number is 20, in the other (D. longifolia) 10; while the hybrid itself has a somatic number of 30. The reduced number in the hybrid, however, is not 15 but 20. Of these 10 are large and 10 small, the latter presumably representing the supernumerary, and hence unpaired, chromosomes of the D. rotundifolia parent.
[54] In their 1905 paper J. B. Farmer and J. E. S. Moore describe two successive synaptic stages (e.g. Elasmobranchs), the first during the contraction of the spireme thread, the second during the looping up of the bivalent segments. (In this paper the authors suggest the term “Meiosis” or “Meiotic phase” for the nuclear changes accompanying the two maturation divisions in plants and animals (μείωσις, reduction).
[55] Whitman, Jour. Morph., 1903.