LKSSON 13.! 



ABORTIVE ORGANS. 



95 



very interesting to the botanist, since it completes the symmetry of 

 the blossom. And to show that this really is the lost stamen, it 

 now and then bears an anther, or the rudiment of one. So the 

 flower of Catalpa should likewise have five stamens ; but we seldom 

 find more than two good ones. Still we 194 



may generally discern the three others, 

 as vestiges or half-obliterated stamens 

 (Fig. 196). In separated flowers the 

 rudiments of pistils are often found in 

 the sterile blossom, and rudimentary sta- 

 mens in the fertile blossom, as in Moon- 

 seed (Fig. 177). 



250. Multiplication of Parts, Quite in 



the opposite way, the' simple plan of the 

 flower is often more or less obscured by | - 

 an increase in the number of parts. In 

 the White Water-Lily, and in many 

 Cactus-flowers (Fig. 197), all the parts 

 are very numerous, so that it is hard 

 to say upon what number the blos- 

 som is constructed. But more com- 

 morlv some of the sets are few and 

 definite in the number of their parts. 

 The Buttercup, for instance, has five 

 sepals and five petals, but many sta- 

 mens and pistils ; so it is built upon 

 the plan of five. The flowers of Mag- 

 nolia have indefinitely numerous stamens 

 and pistils, and rather numerous floral 

 envelopes ; but these latter are plainly distinguishable into sets o 

 ;hree ; namely, there are three sepals, and six petals in two circles 

 or nine in three circles, showing that these blossoms are con- 

 structed on the number three. 



FIG. 194. Corolla of a purple Gerardia laid open, showing the four stamens ; tho cross 

 shows where the fifth stamen would be, if present. 



FIG. 195. Corolla, laid open, and stamens of Pentstemon grandiflorus of Iowa, &c., with 

 a sterile filament in the place of the fifth stamen, and representing it. 



FIG. 196. Corolla of Catalpa laid open, displaying two good stamens and three abortiv* 

 vestiges of stamens 



