1887.] NATUEAL SCIENCES OF PHILADELPHIA. 329 



this point from the want of correspondence in number (abortion ?) 

 of some of the parts; but it is worthy of note that in the two genera 

 just named, and others, where these teeth exist, a tooth is always 

 alternate with the main "lobe" that distinguishes the ray from the 

 tubular floret. 



I think it will be conceded that, given a very little arrestation, de- 

 velopment, union or separation of parts as we find them in Cephal- 

 antkus, it would not be difficult to construct a Composite flower. 



Amorpha canescens, Nutt. Amorplm fruticosa, has leaves gland- 

 ular pellucid punctate, the dots being irregular in size. The legumes 

 are covered by very large glands, also irregular in size. In the 

 leaves of Amorpha canescens the pellucid dots are also irregular, and 

 in many leaflets so indistinct that one should scarcely note their ex- 

 istence, only that we are looking for them. In the legumes, however, 

 the glands are large and abundant, and readily observed through 

 the dense mass of woolly hair that envelopes them. The point is 

 worth noting. The carpels, morphologically leaves, usually have 

 many characters suppressed in the progress of development from the 

 primordial leaf to the carpel ; but here we find characters existing in 

 a marked degree in the leaves of an allied species, and nearly sup- 

 pressed in its own foliage, reappearing as a strong feature in the 

 legume. 



During anthesis many interesting points present themselves. The 

 anthers are fully formed and exhibit their yellow surfaces through 

 the bursting sepals, when they seem to have their further growth 

 arrested, and the pistil only elongates. It extends to double the 

 length of the calyx. This is the work of the first day of opening. 

 The next day the stamens lengthen, but one at a time. As soon as 

 the first one has j-eached the exact length of the ^Distil, another starts 

 into growth, then another, until towards the end of the day's work 

 all are of about equal length. Soon after reaching its final growth 

 the anther sacs burst, and thus one after another, in regular succes- 

 sion, the pollen is ejected from the cells. On the third day the 

 pistil recurves, the apex forming a right angle with the base, and 

 usually having its stigmatic apex in among the withered stamens of 

 the flowers lower on the spike. All this time, the only petal pos- 

 sessed by the flower, the vexillum, has remained nearly quiescent. 

 It projects its beautiful blue tint just a little beyond the calyx, the 

 second day, after the last stamen has made its growth ; but it is not 

 till the third day that it makes a growth in earnest, when it goes on 



22 



