EMBRYOGENESIS IN FLOWERING PLANTS 299 



be used, while in some strains of Mangifera indica (the mango: 

 Anacardiaceae) it affords the readiest means of propagating desirable 

 varieties unchanged; zygotic embryos, on the other hand, yield a wide 

 range of fruiting types. Because of their importance in horticulture, 

 many attempts have been made, thus far with little success, to induce 

 adventive embryos at will. Haberlandt (1921, 1922) tried to achieve 

 this in Oenothera by pricking and squeezing the ovules, on the hypothesis 

 that the proliferation of the embryo-forming cells might be due to a 

 necro-hormone proceeding from adjacent degenerating cells. In one 

 ovule he obtained two embryos which he considered to be of nucellar 

 origin, but this is disputed by Beth (1938) who repeated the experiment 

 in Oenothera and other plants without success. 



Blakeslee (1941) has observed that haploid plants of Datura 

 stramomum, which apparently originate from unfertilised eggs, are of 

 frequent occurrence. This suggested that it should be possible to 

 produce such plants by chemical or other treatment, but so far positive 

 results have not been obtained (van Overbeek and Conklin, 1941). It 

 has become apparent, however, that successive chemical stimuli are 

 involved in embryo formation. These include a stimulus from the grow- 

 ing pollen tubes which initiates development in the placentae and 

 integuments, a stimulus for the further growth of the ovary which 

 prevents its abscission, and a stimulus for the development and differen- 

 tiation of the unfertilised eggs. 



Overbeek, Conklin and Blakeslee (1941) have explored the possi- 

 bility of inducing parthenogenesis in the egg cell of Datura stramomum 

 by injecting the ovary with various chemical substances. These attempts 

 were unsuccessful, but when a 0-1 per cent solution or emulsion of the 

 ammonium salt of naphthaleneacetic acid or indolebutyric acid was 

 injected, several curous multicellular warty outgrowths of what appeared 

 to be the integumentary tapetal layer eventually filled the embryo sac. 

 These undifferentiated tissue masses, which had typically diploid nuclei, 

 and which had originated in the same manner that adventive embryos 

 originate from the tapetal layer, are described as 'pseudoembryos.' In 

 experimental studies of Hosta, in which it has long been known that 

 fertilisation is essential for the production of adventive embryos, 

 Fagerlind (1946) found (i) that pistils receiving abundant pollen of the 

 same species set seeds normally, the embryos being of nucellar origin; 

 (ii) that pistils receiving an inadequate supply of pollen yield some 

 normal and some small ovules, the latter showing neither pollen tubes 

 nor endosperm formation and only incipient adventive embryogeny; 

 (iii) that various foreign pollens had no inductive effect; and (iv) that 

 pistils treated with 1 per cent indoleacetic acid in lanoline remained 

 attached to the plant and three weeks later the ovules contained young 



