oj Gcutiana cauipestria. 891 



structure. Professor lleuslow lias rcx'ordcd instances iu the 

 inif^iionette of the ovules being transformed into leaves, either 

 solitary or rolled round an axis, of which the nucleus is the ter- 

 niiiKition. Others have jiiaiiitained that in certain families the 

 ovules arc parts of tJie carpellary leaves themselves. Jn confir- 

 mation of this, jM, Ad. Bronguiart has j)ublished an account of a 

 njonstrosity of Delphhiium elatum. On the borders of the car|)eLs 

 were observed all states of transition from lateral tritid lobes of 

 the leaf to true ovules. The lateral teeth of these lobules became 

 atrophied ; the middle part was hollowed and cui'vcd ujjwavdsand 

 inwards in the form of a hood, so as to constitute the primiue. 

 The nucleus was descriljcd as originating from a cellular excres- 

 cence or papilla situated on the upper surface, upon the median 

 ncne of each lo])e a little below its summit. M. Brongniai't con- 

 siders it a new production, a cellular pai)illa, developed on the 

 superior face of the middle lobe of the leatlet, and iu the cavity 

 which the latter had formed. Some of the ovules observed in 

 the gentian appeared at first to confirm this oj)iuion. In exa- 

 mining its merits, however, it ap])eai-s necessaiy to take into ac- 

 count two cii'cumstances, the order of development of the ovules, 

 and that of their individual parts. From careful examinations 

 made some years ago, I have been convinced that in some carpels 

 whose ovules are numerous, the order of development is from the 

 base to the apex. In very early stages of the carjiel, the ovules are 

 confined to the lower part alone, there being no trace of them to- 

 ward the uj)pcr portion of the placenta. At a more ad\'anccd 

 stage they occur through the greater ]iart of its extent, but still 

 there is a very evident difference (previous to impregnation) be- 

 tween the progress made by ovules from the base of a placenta^ 

 and those nearer to its apex ; this is obvious to the unassisted 

 eye in regard to the development of the membranes, but actual 

 mcasm*emcnt removes all doubt. The Viola canina may be men- 

 tioned as an example within the reach of those who choose to 

 investigate this matter. In regard to the second point, viz. the 

 order of development of the individual parts (reference is here 

 made to the three outer membranes only), it is unnecessary to say 

 nmch. In the young capsules of the violet ah-eady alluded to, 

 it will l)e found that each ovide first appears toward the lower 

 part of the placenta as a cellular papilla, the nucleus ; the secun- 

 dine soon makes its appearance as a cellular ring around the base 

 of the former ; and lastly, another ring, the rudiment of the pri- 

 miue, appears on the outside of the secundine ; the outer mem- 

 brane, however, becomes rapidly developed, inclosing and con- 

 cealing the other two. It may be also worthy of notice, that the 

 ovules make thcii* appearance while the carpel is still entirely 

 cellular, there being no trace of any va,scular tissue till a more 



