Cleave7's, 1 1 3 



destructive insects from creeping up the petiole. But 

 in all the stellate plants the two little stipules on each 

 side of each leaf have grown gradually out into active 

 green foliar organs, to supplement and assist the 

 leaves, until at last they have become as long and 

 broad as the original leaflets, and have formed with 

 them a perfect whorl of six or eight precisely similar 

 blades. How do we know that .'' you ask. In this 

 simple way, my dear sir. The other Rubiaceae — that 

 is to say, the remainder of the great family to which 

 the stellate tribe belongs — have no whorls, but only 

 two opposite leaves ; and we have many reasons for 

 supposing that they represent the simpler and more 

 primitive type, from which the stellate plants are 

 specialised and highly developed descendants. But 

 between the opposite leaves grow a pair of small 

 stipules, occupying just the same place as the whorled 

 leaflets in the goose-grass ; and in some intermediate 

 species these stipules have begun to grow^ out into 

 expanded green blades, thus preserving for us an 

 early stage on the road towards the development of 

 the true stell^tes. Accordingly, we are justified in 

 believing that in the whorls of goose-grass the same 

 process has been carried a step further, till leaves and 

 stipules have at last become almost indistinguishable. 

 There is, however, one way in which we can still 



