Organization in Phcenogamous Plants. 189 



or through the reflexion above described {ovulum anatropum). 

 I should however be far exceeding the limits assigned me 

 were I to insert here a detailed account of the numberless in- 

 dividual peculiarities which I have met with in the course of 

 my observations. I will content myself with remarking that 

 the Quartine of Mirbel does not exist : what he describes is 

 nothing else than a temporary endosperm in those families, 

 in which the embryo sac displaces the entire nucleus at an 

 early period, although it is not destined to form albumen at a 

 later period by means of a permanent endosperm. 



These integuments experience manifold changes during the 

 ripening of the seed, so that the original number can seldom 

 or never be recognised in the ripe seed. Sometimes all the 

 integuments become consolidated so as to form but one ; at 

 other times, and this is more frequently the case, the integu- 

 ments become separated into different layers of cellular tissue, 

 of various degrees of development, in which case the homo- 

 geneous tissue can easily be separated from the heterogeneous. 

 In this manner the integument of the ripe seed may sometimes 

 be divided into as many as five layers, although only one or 

 two membranes, or, as in Canna^ no complete integuments were 

 originally present. But since it frequently happens that the 

 greatest variety may occur in the ripe seed in this particular 

 in one and the same family*, as has been already related of 

 the group Menyanthece ; whilst, on the other hand, the entire 

 absence, or the presence of one or two integuments in the 

 ovule appears to be very constant in the different families 

 and groups, it may possibly be more advantageous to return 

 entirely to the old terminology of Richard, and only speak of 

 an episperm in the ripe seed, the different positions of which 

 may then be more minutely characterized, whilst at the same 

 time greater accuracy maybe observed in the description of the 

 ovule. Many interesting results may probably be ascertained 

 when these investigations shall have been extended over all 

 the families ; already the small circle of my own observations 

 has afforded many hints. It is remarkable, for instance, that 

 not a single monocotyledonous family possesses fewer than two 

 integuments, and the first impression caused by a review of 

 the different families given above is, that amongst the dicoty- 

 ledonous the majority of the monopetalous families is furnished 

 with but one integument, whilst the polypetalous generally 

 possess two. 



* Indeed in the same genus. Thus one portion of the Salvia; possesses 

 spiral cells in the epidermis of the integument of the seed, and in the re- 

 mainder they are absent. 



[To be continued.] 



