[26] For outline of fertilization see article [Reproduction].
[27] e.g. lymph and various epithelial and connective tissue cells of salamander larva (Flemming, 1891; Heidenhain, 1892); pigment cells of fishes (Solger, 1891); red blood corpuscles (Heidenhain, Eisen, 1897); and numerous other cases.
[28] For an interesting development of this subject see Watasé (1894). This author not only identifies the centrosome with the structures seen in lymph cells, &c., but compares it to the basal granules of ciliated cells and to the varicose swellings on the sarcostyles of striped muscle cells!
[29] The force of this evidence is admitted by Boveri himself. Meves, however, maintains the possibility that the numerous centrosomes appearing in the egg arise by the rapid fragmentation of a centrosome already present.
[30] Cf. especially the behaviour of the centrosomes in the fertilization of the egg of Pleurophyllidia (MacFarland, 1897) and that of Cerebratulus (Coe, 1901). Not only may the sperm centrosomes totally disappear before reaching the egg-nucleus, but in the latter type the definitive centrosomes appear while the last traces of the sperm asters are still visible.
[31] e.g. Meves; Spermatagonia of Salamandra.
[32] Cf. especially the artificial production of amitosis in Spirogyra; W. Pfeffer, 1899.
[33] Cf. Boveri, 1904, p. 13. (For Boveri’s criticism of Delage’s views, cf. Boveri, 1901 and 1902.)
[34] It should, however, be noted that the assumption that a particular group of characters remains always associated in a particular chromosome is one that is very difficult to reconcile with the mode of inheritance of Mendelian pairs of characters in the case of organisms with a relatively small chromosome number.
[35] Boveri (1902), “Fertilization of enucleated Echinus-egg fragments,” and M. Boveri (1903); by shaking the egg shortly after fertilization the sperm centrosome is prevented from dividing, and a monaster instead of a diaster results, the divided chromosomes remaining in the one nucleus.