UMBELLIFERJE, 87 



At maturity, the fruit, enveloped in the receptacle to which it adheres, 

 is a d'uiclicnc finally separating into its two component achenes,* 

 between which is a slender column (columella or carpophonmi ^) often 

 divided in its upper part into two branches, each rising above the 

 middle line of the ventral face of the achene. One of the achenes, 

 the anterior, is surmounted by the two very short but persistent 

 anterior sepals ; the other, the posterior, bears three. ^ Vertical 

 nervures traverse the entire length of the back of each achene. 

 There are first five primary ridges,^ then a middle dorsal, two marginal 

 which correspond to the margins of the achene, and two lateral, 

 intermediate between these and the middle one. These ridges are not 

 very prominent, and their free margin is charged with small superposed 

 prominences, surmounted by a variable number of small hairs. Be- 

 tween these nervures are four secondary ridges,^ which in the Carrot 

 are more developed than the primary. Their free margin is divided 

 into conical points, ranged in a single vertical line. The secondary 

 ridges are inserted in their entire length in the vertical furrows of the 

 achene called valleculce. These contain linear reservoirs of coloured 

 and odorous gum-resinous juice called vittce.^ The surface by which 

 the two achenes are united is named the commissure.'^ This may also 

 bear two bandlets, one on each side of the middle line. In the interval 

 between the two commissural vittse, the surface of the achene is fiat 

 or very slightly convex, like the internal surface of the seed on which 

 the pericarp appears to rub. In each achene is 'a descending seed, 



1 Or mericarT^s {mencarpia) . They are formed sepals on one meriearp as on the other, since 

 for the most part of the concave receptacle, the they have each the same number of primary 

 part representing the carpellary leaves here ridges. The surface of carpellar separation, 

 occupying only the upper portion of the fruit. near the insertion of the sepals, is not flat but 



2 Columella, carpophorum. It is often undi- undulated. M. ue Lanessan {Bull. Soc. Linn. 

 vided in the genus Daucus as in many others. Par. 17, 23) has much better observed the com- 

 DucHARTRE {op. cU. 1129) is therefore wrong in portment of the sepals and ridges at the time 

 giving as a general character of the TJmhelliferm of dehiscence, as the disjunction of the carpels 

 " mericarps suspended by a bipartite filament." in the Umbelliferce, themselves indehiscent, is 

 Genera have even been founded on the divided often called. It is the receptacle enclosing the 

 or undivided character of the carpophor (see, carpels which separates into two halves. 



on the nature of the carpophor, H, Mohl, But. ^ Juga primaria or carenalia. 



Zeit. (1863) 264 ; — H. Bx. Bttll. Soc. Linn. Par. ^ Jaffa secundaria or stdnralia. 



161). ^ These reservoirs belong to the category o 



3 From this is evident the error of those who secreting canals. On their constitution, se 

 suppose that the primary ridges correspond to Trkcul, Jnn. Sc. Nat. ser. 5, v. 275. 



the middle nervures of the sepals. If this ■ ' Or commissural, ventral face, 

 theory were admissible, there would be as many 



