w « 



ry 



iNTRODUCTION. . XVll 



the 2)crfcct pistil there is always at least one ovule in the ovary, and some portion of 

 shginatic surface. Witliout these the pistil is imperfect, and said to be barren (not 

 setting seed), aloriive, or rudimentary (84), according to the degree of imperfection. 



123. The ovary being tlie essential part of the pistil, most of the terms relating to 

 the number, arrangement, etc., of the carpels, apply specially to their ovaries. In some 

 works each separate carpel is called a pistil, all those of a flower constituting together 

 the gyncecium; but this term is in little use, and the word pistil is more generally 

 applied in a collective sense. Vriien the ovaries are at all united, they are commonly 

 termed collectively a compound ovary. 



124. The number of carpels or ovaries in a flowsr is frequently reduced beiow that 

 01 tlie parts of the other floral whorls, even in flowers otherwise symmetrical. In a 

 Tery few genera, however, the ovaries are more numerous than the petals, or indefinite, 

 J hey are in that case either arranged in a single whorl, or form a head or spike in the 

 centre of the flower. 



125. The terms monogifuous^ d igy nous, polygy nous y etc. (with a pistil of one, two, or 

 n^ore parts), are vaguely used, applying sometimes to the whole pistil, sometimes to the 

 ovaries alone, or to the styles or stigmas only. Where a more precise nomenclatm-e is 

 adopted, the flower is 



monocarpellary, when the pistil consists of a single simple carpeL 

 ^ 6/-, tri-^ etc., to poly-carpellary, when the pistil consists of two, three, or an inde- 

 finite number of carpels, whether separate or united. 



syncarpous^ when the carpels or their ovaries are more or less united into one 

 compaund ovary. 



apocarpous, wlien the car|)els or ovaries are all free and distinct. 

 12G. A compound ovary is 



unilocular or one-celled, when there are no partitions between the ovule?, or when 

 ^ncsQ partitions do not meet in the centre so as to divide tlie cavity hito several cells. 



pluriloctilar or several-celled, when completely divided into two or more ceUs by 

 partitions called dissepiments {sepia), usually vertical and radiating Irom the centre or 

 axis of the ovary to its circumference. 



bi', tri-^ etc., to miiUi-lociiIar, according to the moiibcr of these cells, two, three, 

 ^tc, or many. 



127. In general the number of cells or of dissepiments, complete or partial, or of 

 rows of ovules, corresponds with that of the carpels, of which the pistil is composed, 

 -l^ut sometimes each carpel is divided completely or partially into two cells, or has two 

 I'pws of ovules, so tliat the number of carpels appears double what it really is. Some- 

 tmes again the carpeb are so completely combined and reduced as to form a suiglc ceU, 

 ^■ith a single ovule, although it really consist of several carpels. But in these c^s the 

 o-^ary is usually described as it appears, as weU as such as it is theoretically suppo<=ed to be. 



128. In apocarpous pistils the styles arc usually free, each bearing its own stigma. 

 V ery rarely the greater part of the styles, or the stigmas alone, are united, whdst the 

 ovaries remain distinct. 



-l*.J. Synearpous flowers ai'e said to have 



several styles, when the styles are free from the base. 



one style, ivith several hrancfies, \;hen the styles are connected at the base, but 

 separate below the point where the stigmas or stigmatic surfaces commence. 



one simjile style, ny't/i several sttymas, when united up to tlie pomt where the 

 s Jgmas or stigmatic surfaces commence, and then separating. 



one simple style, with a branched, lobed, toothed, notched, or entire stiyma (as the 

 oase may be), ^hen the stigmiis also are more or less united. In many works, bow- 

 er, this precise nomenclature is not strictly adhered to, and considerable confusion is 

 olten the result. 



130. In general the number of styles, or branches of the style or sti.gmn, is the same 

 ^ that of the carjiels, but sometimes that number is doubled, especially m the stigmn^ 

 s-Tia sometimes the stigmas are dicliotomously or pinuately branclied, or i^f/"t*^^/« ^, 

 Jl^at 18, divided into a tuft of hair-like branches. All these variations sometimes make 

 t a difficult task to determine the number of carpels forming a compound ovary, but 

 ^^'^ point is of considerable importance in fixing the affinities of plants, and, by careful 



ev 



