( 272 ) 



tetrad. In the pollen mother-cells also tlu divisions lake place regularly, 

 but after 4 tetrad cells have been formed, only two cells, in general, 

 develop properly, the other I wo only partially, so that from them 

 grains aie formed, from which the contents gradually disappear, 

 although the wall is fairly normal in structure. 



Since the divisions arc normal, the appearance of this sterility 

 need not at all he the result of a hybrid nature. It follows from 

 the literature on sterility, thai, besides in hybrids, sterility occurs in 

 many other plants, and thai in most hybrids, which have been 

 examined cytologically, sterility was already present in one or both 

 parents. And since an examination of one hundred species showed 

 sterility to be pretty common among' Onagraceae, we are not, in 

 my opinion, justified in deducing from the sterility of a plant that 

 it is a hybrid. If the above-mentioned objection of Bateson against 

 the mutation-theory is to have value, it will be incumbent upon 

 him to name the presumptive parents of Oen. Lam. 



In mature ovaries those ovules, which are not destined to develop, 

 may be recognized by their more transparent nucellus, not containing 

 an embryosac. In such ovulus the penetration of a pollen-tube was 

 never observed, whereas this was repeatedly found in the normal 

 ovules of the same section. // would appear, therefore, that the 

 normal embryosac ('.veils an attraction on the pollen-tubes. 



\\\ combining the results of the investigation of the ontogenetic, 

 with those of the cytological development, it appears, that Oen. Lam. 

 is very suitable for the experiments, referred to in the introduction, 

 because the cytological development of the pollen and of the ovules 

 are sharply separated in point of time. In flowers of 'M) nun. the 

 development of the pollen is almost complete, while in the same 

 flowers the development of the mother-cells in the ovule is only 

 beginning. It would therefore be possible to influence the pollen 

 and the ovules separately, and the flowers which should be selected 

 for this purpose, can be easily recognized after a little practice. 

 Should one wish to influence the pollen before the synapsis, then 

 llowers 10 — 11 m.ni. long should be selected, in which the stamens 

 have a length of 3 mm., since the synapsis of the pollen takes place 

 in llowers of 12 — 13 mm., in which the anther and also the filament 

 have a length of 4 mm. The synapsis of the embryosac mother-cell 

 lakes place in flowers of about 3V a cm. In order to influence this, 

 flowers of 3 — 3 1 /, cm. should therefore be selected, in which the 

 ovary and the calyx tube have about the same length i.e. 4 1 /, ram, 



1 hope soon to be able to publish a more detailed description, 

 with plates, of the results of I his invest igal ion in the Reciieil des 

 Travaux Botaniques Néerlandais. 



