rHASEOLUS DIVERSIFOLIUS. KIDN'EV-BEAN. I7I 



leaves lobed and some entire, and others again had all of them 

 entire. The flowering specimen on our plate was drawn from 

 one of these entire-leaved plants, while the lolled leaf in tlie 

 illustration was taken from another plant in whicli nearly all the 

 leaves differed. 



It is exceedingly interesting and important to observe such 

 variations, as they have much influence in establishing the doc- 

 trine of evolution, since they tend to show how one form may 

 grow out of another. Thus the leaflets represented in Fig. 2 

 on our plate are not only lobed, but a strong midrib has also 

 been formed in each lobe. These midribs have their separate 

 lateral veins, and we can easily see that, with a little more effort 

 there would be a complete subdivision, making a leaf of nine 

 leaflets instead of three, as now. Again, we may learn, from 

 the lessons afforded by our plant, how one species may event- 

 ually branch out into several distinct species. As matters stand 

 at present, we cannot base different species on the different 

 shapes of the leaves, because these several shapes are variable, 

 and may occur now on this, now on that plant, or perhaps even 

 all together on one and the same plant. But if it should ever 

 happen that the descendants of certain plants, under a variety 

 of circumstances, and through a long series of years, have only 

 entire leaves, while the descendants of others have only lobed 

 leaves, without cither of them showing any sign of a desire to 

 return to the variable condition, we should then be justified in 

 looking upon the shape of the leaves as a fixed character, and 

 in building a species upon this character, although we knew 

 that all these plants originally came from one stock. In further 

 illustration of what we have said, we may point to P. lu-lvolus, 

 a species allied to P. divcrsifolius, and probably of the same 

 origin. One of the chief characters at present relied on to dis- 

 tinguish these two species is the length of the flower-stalks, 

 w^hich latter in P. divcrsifoliiis are always much shorter in pro- 

 portion to the leaves than in P. Jicholus. 



Our plant trails over the ground in its wild state, but under 



