370 METAMORPHOSIS 



extract of sea-urchin sperms, containing no nuclein, and as a consequence of this 

 treatment the ovum developed. 



We thus arrive at the conclusion that the egg needs a developmental 

 stimulus before commencing to divide and grow (compare Solms, 1900, and 

 the literature cited there), and such developmental stimuli we have already 

 found to be very widely distributed. Thus we have seen that buds whose 

 developmental activity has been temporarily inhibited may be made to unfold 

 by ether, that the spores of many mosses germinate in the dark only when the 

 temperature is raised, and that high temperatures are quite generally essential 

 in propagation by cuttings. It is also known that the pollen-tubes in many 

 cases induce development, apparently not by fertilization but by the secretion 

 of some soluble material ; thus the seed initials in Orchidaceae develop only if 

 pollen-grains germinate on the stigma. [The same is true of Fntillaria -persica, 

 according to Strasburger (1886).] And this stimulus, doubtless chemical in 

 its nature, may, according to Treub (1882), also be effected by certain insects 

 in one of the tropical orchids. In addition to this, the germination of the 

 pollen-tube has an exciting influence in the development of the fruit. This is 

 particularly noticeable in certain cultivated plants, which, as for example 

 currants and Sultana raisins, produce no seeds, the ovules having degenerated. 

 If the stigmas of these plants be not pollinated, the fruit faUs to develop, 

 but pollination causes development without leading to any fertilization (Muller- 

 Thurgau, 1898, compare also Noll, 1902 [Massart, 1902J). 



Sufficient examples of the renewal of development as a result of stimulus 

 action have now been given to show that the special effect of sperms is not 

 without analogies. This is probably the most appropriate place to discuss certain 

 phenomena which appear in the embryo-sac. We have comparatively recently 

 learned that in Phanerogams not only does one sperm-cell from the pollen-grain 

 fuse with the ovum, but that the second sperm-cell also passes into the embryo- 

 sac and fuses with the united polar nuclei before these give rise to the endo- 

 sperm (Nawaschin, 1898 ; Guignard, 1899). As to the phylogenetic meaning 

 of this second act of fertilization we need not here speak. We are only concerned 

 at present with the fact that we appear here to be dealing with the removal 

 of a developmental check. Apparently all the nuclei of the embryo-sac are 

 incapable of development without some developmental stimulus, but the requisite 

 stimulus is not always a nuclear fusion. There is a whole series of plants in 

 which the embryo, as in apogamous ferns, arises not from the ovum but from 

 neighbouring cells (compare Ernst, 1901), it may be from the synergidae, the 

 antipodal cells, or from cells of the sporangium wall external to the spores — i. e. 

 from the nucellus. In the last case especially a fusion with a male cell is entirely 

 out of the question. But a definite external stimulus is in many cases necessary 

 for the formation of such adventitious embryos. In Nothoscordon fragrans, for 

 instance, the adventitious embryos first appear only after the ovum is normally 

 fertilized. After recent discoveries a reinvestigation of the question is much 

 to be desired, all the more as in certain other cases where adventitious embryos 

 occur, e.g. Coelohogyne ilicifolia (Strasburger, 1878) and probably also Eu- 

 phorbia dulcis (Hegelmaier, 1901), they certainly arise without the previous 

 action of pollen-tubes or sperm-cells. All stages, from normal embryo forma- 

 tion to the complete parthenogenesis of Antennaria and Alchemilla, and the 

 adventitious embryo formation in Coelohogyne, occur in nature. In the last- 

 mentioned cases we must look for an internal stimulus which initiates the de- 

 velopment of the cells concerned. 



The above examples of adventitious embryos are also of interest from 

 another point of view. It appears that all cells contained in the embryo-sac or 

 which come to lie therein, assume a form similar to the normal embryo ; the 

 embryo-sac cell must be able to exert a similar stimulus (Juel, 1900 a ; 

 Strasburger, 1878). 



We may now return to the phenomena of normal fertilization. We have 



