PARTHENOGENESIS 369 



striae. These granules are capable of division, and one is found temporarily 

 at each pole of the nuclear spindle. The idea is now very generally accepted that 

 these centrosomes play an important part in nuclear division, since they deter- 

 mine the arrangement of the nuclear spindle and fix the dynamic centres, at least 

 certain observations of BOVERI may be interpreted in this way. If then the egg 

 forms no centrosome at all, or no active centrosome, then we can easily understand 

 how its power of division is inhibited and how perhaps it cannot even grow. But 

 BOVERI (1902) has not taken into consideration the fact that his theory will not 

 hold in plants, where centrosomes are absent. If we follow STRASBURGER 

 (1900 a) in assuming that in the higher plants the centrosome may be represented 

 by a partly-defined portion of the protoplasm, the so-called kinoplasm, we may 

 extend BOVERI'S hypothesis to the plant world also. 



One thing is certain, viz. that none of the theories mentioned above can 

 explain all the facts. All three agree in supposing that something is wanting 

 in the egg, and so far all these hypotheses are correct. The phenomena of partheno- 

 genesis however prove that this conception is not universally correct. True 

 parthenogenesis, that is to say, the development of an unfertilized egg into an 

 embryo, is not often met with in plants. Where it has been carefully investigated, 

 as in Antennaria alpina (JuEL, 1900 a), and in species of Alchemilla (MURBECK, 

 1901), it has evidently become the normal process for the propagation of the 

 species, and it is questionable whether the ovum in these plants is capable of being 

 fertilized at all. The question would well repay an investigation, for if normal 

 pollen does not occur in the species in question the experiment should be made 

 with that of another closely-allied species. The investigation should prove of 

 special interest, since JUEL has found that in Antennaria the egg contains the 

 full complement of chromosomes present in the somatic cells, and he holds, for 

 this reason, that fertilization is impossible. If KERNER'S belief that Antennaria 

 hansii, Kern., is a hybrid between A. alpina $ and A. dioica a* be well founded, 

 then JUEL'S position will be untenable (compare FOCKE, 1881, 194). [Mean- 

 while STRASBURGER (1904^ has shown in the case of Alchemilla and JuEL(i9O4) 

 in the case of Taraxacum, both of which are parthenogenetic, that no reduction 

 takes place in the formation of the ovum. The frequency in change of view 

 on this subject and the contradictory nature of the observations make it neces- 

 sary to proceed with caution. The fact established by OSTENFELD (1904) that 

 certain species of Hieracium can produce embryos both sexually and partheno- 

 genetically is of the greatest interest. Since it is unlikely that there are in 

 these plants two kinds of ova we are bound to assume that the normal ova 

 in the higher plants may develop parthenogenetically just as easily as those 

 of Marsilia, and that the number of the chromosomes has not that significance 

 which is attributed to it by the cytologists.] So far as the question that princi- 

 pally concerns us here is concerned these two plants are unimportant. Impor- 

 tant on account of their bearing on the present question are certain species of 

 Marsilia of which more than one shows a tendency to parthenogenesis. As 

 NATHANSOHN (1900 a) has shown, it is possible, by raising the temperature, to 

 increase very considerably the percentage of unfertilized eggs which develop. 

 In one investigation on Marsilia vestita, at 18 C. only 1-3 per cent, developed 

 parthenogenetically, at 35 C. 7-3 per cent, developed. The egg has the capacity, 

 whether fertilized or unfertilized, of developing into an embryo, and in the 

 latter case it is not the entry of a material substance but a rise in temperature 

 that supplies the requisite stimulus. We must realize therefore that the 

 stimulus may be supplied by a rise of temperature, just as by material contained 

 in the sperm. That this material need not necessarily be nuclein or an organ- 

 ized portion of the cell, such as a centrosome or chromosome, has been shown 

 by the experiments of LOEB (1899-1902) and WINKLER (1900) on animals. 

 The ovum of the sea-urchin, which in nature only develops after fertilization, 

 was treated by LOEB with magnesium chloride, and by WINKLER with a watery 



JOST B b 



