272 DE MERCKLIN'S OBSERVATIONS. [BOOK i. 



tripennatiparted, &c. Finally, in cases where leaves are 

 extremely divided, and the parenchym of the ultimate ramifi- 

 cations of the veins does not unite and form lobes, we say, 

 in general terms, that the leaf is multifid, laciniated, decom- 

 posed, or slashed; terms which express the appearance of a 

 leaf, without any very precise signification." 



With regard to compound leaves, their leaflets always have 

 the primary veins running at an angle more or less acute 

 towards the margin. "This is perfectly intelligible if we 

 reflect that their lateral veins represent not the primary, but 

 the secondary and tertiary veins of simple leaves, which latter 

 are always pennated. 



" The leaflets of pennated leaves are usually placed opposite 

 each other in pairs along a common petiole. These pairs of 

 leaves are called in Latin jug a : thus a leaf with one pair is 

 unijugum ; with two pairs, bijugum, &c. 



" Usually one of the leaflets terminates the petiole ; the 

 leaf is then unequally pinnated (imparipinnatum) ; but some- 

 times there is no odd leaflet, and the petiole ends abruptly, or 

 in a point or tendril: this is equally pinnated, (pari-pinnatum.) 



" Sometimes the leaflets themselves are subdivided (folium 

 bipinnatum, tripinnatum) . In this case the lateral petioles 

 which bear the leaflets are called partial ; and the small sup- 

 ports of the leaflets themselves, stalklets (petiolules)." 



Such are De Candolle's ideas of the typical formation of 

 leaves. They offer a convenient mode of studying the modi- 

 fications in structure of these organs, but they are by no 

 means confirmed by the accurate investigation of De Mercklin, 

 who gives the following as the result of his inquiry. 



" First Period. Birth of the Leaf considered generally. 

 On stripping a leaf-bud of its leaves, the latter are found to 

 decrease in size in proportion as you approach the centre, and 

 at last a microscope is necessary to distinguish them. The 

 bud ends in a round or conical projection, which has no 

 proper envelope. This projection is called the summit of the 

 axis (Punctum vegetationis of Wolff) ; its lower extremity is 

 the nucleus of the bud. In some natural orders, as in Cucur- 

 bits and Crassulacese, the summit of the axis is reduced to 



