88 DISSECTION AND ANALYSIS OF THE FLOWER 



been so thoroughly considered, that it need not be again taken up except 

 to say that glands, which are frequently metamorphosed stamens or 

 appendages to the several parts, must not be mistaken for a disk. With 

 the flower in this position it may also easily be seen whether the parts 

 are cyclical, and if so the number of circles may be determined. If 

 duplication has occurred, its origin in chorisis or metamorphosis is 

 readily determined, while if suppression has occurred it can readily 

 be referred to the respective circle. 



The relation of the parts to one another having been thus determined, 

 each of them must next be studied individually. The shape and texture, 

 and the division into parts, with the details of any existing appendages, 

 will be sought separately in sepal, petal, and stamen and in filament 

 and anther separately. In the examination of the stamen, it is essen- 

 tial that it be examined separately in direct lateral, ventral, and dorsal 

 views, as only thus can the true relations of its parts become known. 

 The form of attachment of anther to filament and the point of junction 

 between filament and connective are next in order, as well as the form 

 of dehiscence of the thecae and especially the position and direction 

 assumed by the sutures, pores, or valves of the latter. The chief diffi- 

 culty in the examination of the stamens ■wall be in determining the part 

 upon which any existing appendages originate. The position which 

 such an appendage occupies is frequently quite misleading as to the 

 nature of its origin, and it must be carefully moved about \dth the 

 points of the needles, great care being taken that no dehcate attachment 

 is severed, before it can be definitely ascertained whether an appendage 

 originates from filament, connective, or theca. 



It is, moreover, not rarely the case that the characters of appendage 

 and anther are so concealed or -even substituted that the one may be 

 mistaken for the other. The examination of the stamen is not com- 

 plete until the characters of the pollen, as to its being granular or 

 collected into pollinia, the nature and characters of the latter, and even 

 the characteristics of the individual pollen-grain, have been determined 

 by the aid, chiefly, of the compound microscope. 



The gynaecium, still in position upon the torus, must next be studied 

 as to its relations to the latter and its composition of united carpels 

 or separate pistils. If of separate pistils, they must be separately 

 removed from the torus, great care being taken not to mutilate the 

 latter, and their number and regularity must be determined. If 

 regular, the detailed examination of one of them is sufficient, but if 

 irregular, one of each form must be separately studied. 



The external characters of the pistil present no difficulty for exami- 



