1)6 PROCEEDINGS OF THE ACADEMY OF [1899. 



that wliere there should have beeu petals alternate with the sepals 

 and stameus, there are four greenish umbonate glands. These of 

 themselves suggest abortive petals. Holding the flowers up to the 

 light, however, we find by the aid of the lens that the petals are 

 really there, but have been completely covered by the connate 

 sepals, the glands being the thickened apices ! May not the sup- 

 posed stipular gland, in like manner, be some foliar organ almost 

 covered by the union of superior parts ? In aid, farther, of this 

 suggestion, we note that the primary origin of the leaf is far below 

 the point from which the leaf proper or blade seems to spring. 

 The consolidated petiole can be traced far below the base of the 

 expanded portion. Each consolidated petiole, however, grasps only 

 one-fourth of the circumference of the stem, thus giving it a sub- 

 quadrangular character. Why might not the supposed stipule be 

 the apex of another pair of minute scaly bracts, all of which but 

 the extreme j^oints had been covered, as the petals were, by the 

 superior size of the consolidating parts above ? With these sug- 

 gestions we examine the stipular glands closely, only to find that 

 they are opposite to, and not alternate with, the pair of scaly 

 bracts above them. We must still believe they are stipular appen- 

 dages of the leaves, and not the upper remains of independent but 

 submerged organs. 



Of late years it has been found that what appear as striking 

 exceptions to general rule can be accounted for in the regular 

 way, the operation of the law only having been obscured. Why 

 Isnardia should yet seem a striking exception to the absence of 

 stij^ules in Onagrads still remains a problem. 



During the examination of these j)lants the absolute fertility of 

 the flowers was suggestive of perfect self-ferlilization. Few axils 

 but had their primary branching system retarded for the production 

 of a flower instead. Not a solitary imperfect seed-vessel could 

 be found, out of the many hundreds on a single bi-anching plant. 

 Self-fertilization was found to be the fact. Not only do the anthers 

 discharge the pollen simultaneously with the expansion of the 

 sepals; but they rest on the stigma, so that it is impossible for any 

 external agency to interfere with their work. 



Another interesting observation is that at the expansion of the 

 sepals, nectar copiously exudes from the umboues above noted. No 

 Avinged insect appears to visit these minute flowers. 



