186 ORGANOGRAPHY. BOOK I. 



called naked seeds, such as those of Umbelliferse, Labiatae, 

 and Boraginese, and the grain of corn ; but, now that atten- 

 tion has been paid to the gradual development of organs, 

 such errors have been corrected. In cases where a trace of 

 the style cannot be discovered, anatomy will generally show 

 whether a minute body is a seed or fruit, by the presence, in 

 the latter case, of two separable and obviously organically 

 distinct coatings to the nucleus of the seed; but in other 

 cases, where the pericarp and the integuments of the seeds 

 are combined in a single covering, and where no trace of 

 style remains, as sometimes happens, nothing can be deter- 

 mined as to the exact nature of a given body without follow- 

 ing it back in its growth to its young state. This, however, 

 may be stated, that naked seeds, properly so called, are not 

 known to exist in more than three or four orders in the 

 whole vegetable kingdom ; viz. in Coniferse and Cycadeas, 

 where the ovules also are naked, and in Peliosanthes Teta 

 and Leontice, in which the ovules, originally enclosed in an 

 ovary, rupture it at an early period after fertilisation, and 

 subsequently continue naked until they become seeds. 



Such being the case, it follows that all the laws of structure 

 which exist in the ovary are equally to be expected in the 

 fruit ; and this fact renders a repetition in this place of the 

 general laws of formation unnecessary. Nevertheless, as, in 

 the course of the advance of the ovary to maturity, many 

 changes often occur which contribute to conceal the real 

 structure of the fruit, it is in all cases advisable, and in many 

 absolutely necessary, to examine the ovary, in order to be 

 certain of the exact construction of the fruit itself. These 

 changes are caused by the abortion, non-development, obli- 

 teration, addition, or union of parts. Thus the three-celled 

 six-ovuled ovary of the oak and the hazel becomes, by the 

 non-development of two cells and five ovules, a fruit with one 

 seed; the three-celled ovary of the cocoa-nut is converted 

 into a one-celled fruit by the obliteration of two cells and their 

 ovules ; and the two-celled ovary of some Pedalineae becomes 

 many-celled by a division and elongation of the placentae. 



In a very early state the ovary of the Lychnis and of the 

 primrose consists of five cells, each with a placenta having a 



