STRUCTURE OF FLOWERS 719 



purple, or some other bright colour, antlioi-ijn n , erythrophytt, ite.. 

 dissolved in the cell-sap. There are some petals whose cells exhibit 

 very interesting peculiarities, either of form or marking, in addition 

 to their distinctive coloration; such are those of the Pela/rgonium, 

 of which a small portion is represented in fig. 564. The different 

 portions of this petal when it has been dried after stripping it of 

 its epiderm, immersed for an hour or two in oil of turpentine, and 

 then mounted in Canada balsam exhibit a most beautiful variety 

 of vivid coloration, which is seen to exist chiefly in the thickened 

 partitions of the cells ; whilst the surface of each cell presents a 

 very curious opaque spot with numerous diverging prolongations. 

 This method of preparation, however, does not give a true idea of 

 the structure of the cells ; for each of them has a peculiar mam mi 1- 

 lary protuberance, the base of which is surrounded by hairs : and 

 this it is which gives the velvety appearance to the surface of the 

 petal, and which, when altered by drying and compression, occa- 

 sions the peculiar spots represented in fig. 564. Their real character 

 maybe brought into view by Dr. Inmaii's method, which consist^ 

 in drying the petal (when stripped of its epiderm) on a slip of glass. 

 to which it adheres, and then placing on it a little Canada balsam 

 diluted with turpentine, which is to be boiled for an instant over 

 the spirit lamp, after which it is to be covered with a thin glass. 

 The boiling ' blisters' it, but does not remove the colour; and on 

 examination many of the cells will be found showing the mammilla 

 very distinct!}', with a score of hairs surrounding its base, each of 

 these slightly curved, and pointing towards the apex of the mammilla. 

 The petal of the common scarlet pimpernel (Anagallis arvensis), 

 that of the common chickweed (Stellar ia media), together with many 

 others of a small and delicate character, are also very beautiful 

 microscopic objects ; and the two just named are peculiarly favour- 

 able subjects for the examination of the spiral vessels in their natural 

 position. For the ' veins ' which traverse these petals are entirely 

 made up of spiral vessels, none of which individually attain any 

 great length, but one follows or takes the place of another, the 

 conical commencement of each somexvhat overlapping the like termi- 

 nation of its predecessor; and where the 'veins' seem to branch, 

 this does not happen by the bifurcation of a spiral vessel but by 

 the ' splicing on ' (so to speak) of one to the side of another, or of 

 two new vessels diverging from each other to the end of that \vhich 

 formed the principal vein. 



The Anthers and Pollen-grains also present numerous objects of 

 great interest, both to the scientific botanist and to the amateur 

 microscopist. In the first place, they afford a good opportunity of 

 studying that form of ' free-cell-formation ' which seems peculiar to 

 the parts concerned in the reproductive process, and which consists 

 in the development of new cell -walls round a number of isolated 

 masses of protoplasm forming parts of the contents of a parent 

 cell, so that the new cells lie free within its cavity, instead of being 

 formed by its subdivision, as in the ordinary method of multiplica- 

 tion. If the anther be examined by thin sections at an early stage 

 of its development within the young flower-bud, it will be found to 



