234 Miscellaneous. 



which make their appearance occasionally upon various parts of the 

 axis of inflorescence, but most frequently at its base. These bracts, 

 in many species, are like leaves in producing decurrent projecting 

 lines, preserving the characters which distinguish the margins of 

 bracts and leaves. Moreover, when the bracts disappear, they leave 

 traces of their existence in those decurrent lines which render the 

 axis of inflorescence angular. 



The author investigates the cause of the habitual abortion of the 

 bracts, and thinks he has discovered it in the mode of successive 

 development of the leaves and flowers. The stem, in commencing 

 its evolution, is furnished with contiguous leaves; and at the extremity 

 of the cauline axis the rudiments of flowers are produced, which 

 form, not a raceme, but a simple, concave, closely packed corymb, 

 closely enveloped and exceeded by a great number of erect leaves, 

 closely pressed together, which, by their elasticity, present more or 

 less resistance to the expansion of the flowers. The latter are, in 

 fact, developed successively upon a sort of narrow terminal plateau, 

 in which the youngest flowers, placed in the centre, push the older 

 ones from within outwards against the enveloping leaves. It will 

 hence be understood that the bracts, not finding sufficient space for 

 their development, disappear from the midst of the flowers by com- 

 pression, which is so powerful an abortive agent. 



The author also calls attention to the fact that the peduncles are 

 depressed from in front backwards, and the more in proportion as 

 the raceme is fuller and presents a greater resemblance to a corymb, 

 and that the flower-buds themselves present the same fact more or 

 less distinctly, and in the same direction as the peduncles. The 

 anterior and posterior sepals are narrower than the lateral ones, and 

 are never inflated at the base, as if they had been hindered in their 

 development. The petals, having an oblique direction in relation to 

 the sepals, escape in consequence the modifying action of which the 

 efl^ects have just been described, except, perhaps, in some genera, 

 such as Iberis, in which the dissimilarity of the organs seems to be due 

 to the arrangement of the raceme in its earliest stages of development. 



In the theory of an androecium with a double whorl, accepted by 

 the author, the two stamens of the exterior whorl, which are opposite 

 to the upper and lower sepals, are regarded as being habitually 

 aborted, and, in his opinion, in consequence of the same cause which 

 likewise frequently induces the complete abortion of the gland upon 

 which each of these stamens is inserted when they occasionally 

 make their appearance, or allows it to persist, or shortens and deforms 

 it more or less, according to the degree in which its action is exerted. 



He thinks also that if the long stamina, originally distant from 

 each other, subsequently become approximated (usually in two groups), 

 this must be attributed to an oblique pressure, simultaneously taking 

 place in two convergent directions, exerted by the flower-buds upon 

 each other. 



Most authors regard the fruit of the Cruciferae as formed of two 

 carpellary leaves, which seems to be indicated by the number of cells 

 and that of the stigmata. The author , however, following the example 



