138 



5, p. Ill, From flowers which, if not crossed, are self-fertile, it is 

 but a step to such as are regularly self-fertilized, but with a possibil- 

 ity for occasional crossing ; and a step further leads to cleistogamous 

 flowers, incapable of crossing, and constantly self-fertilized. It is a 

 significant fact that the latter are never produced to the exclusion of 

 others capable of being crossed, their occurrence under any circum- 

 stances being merely a sure means of securing some offspring with 

 as little drain as possible upon the system of the plant. 



Through the kindness of Mr. Watson, I have been able to exam- 

 ine the specimens of Scrophidaria in the Gray herbarium, with the 

 following results. So far as could be judged from dry s ecimens, 

 protogyny with the characteristic relations of stamens and pistil 

 found in S, nodosa^ occur in S, aquatica, S. bicolor^ S. calliantha^ S. 



". chrysantha^ S. floribunda^ S. glabrata^ S, grandideutata (?), 

 S. hiria, S. Hispanica^ S. incisa^ S. lacmiata^ S, meTTifera^ S^ 7iiulticaii- 

 hs^ S. O/wer/a/ia, S.pefegrinay S7racejnosa^ Srsanibucifolia^ S, ScopoUi^ 

 S^ Smithii^ S. sylvaiica, S. tricopoda^ S^ tnfoUata^ and S. variegata^ 

 In all of these, crossing appears well provided for, while the posi- 

 tion of the parts in some flowers of several species render it proba- 

 ble that "self- fertilization is possible, later, if this has not occurred. 

 No species were found in which there was reason to believe that 

 synacray and regular self-fertilization occur ; but in a considerable 

 number the specimens did not allow an opinion to be formed. ' Cleis- 

 togene flowers were found only in S. arguta^ where they were first 

 detected by Durieu de Maisonneuve in 1856 (i, p. 569), and later 

 mentioned by Kuhn (i, p. 67) ; an4 it is needless to say that open 

 entomophilous flowers occur on the same plants, although the speci- 

 mens do not allow me to state whether dichogamy is present, or not. 

 Before closing, reference must be made to several conflicting state- 

 ments about the sort of dichogamy found in this genus. As stated 

 before, Sprengel found that the pistil matures before the stamens; 

 this and Euphorbia being the only genera known to him as possess- 

 ing what he called '* weiblich-miinnliche Dichogamie." Axell 

 (^ P- 39) doubts this without having studied the plant; but Hilde- 

 brand (x, p. 21), Muller and others have no difficulty in finding this 

 state of things which we call to-day protogyny, and I know of no 

 statement implying the reverse, saving that by Prof. Beal (r, p. 202), 

 which is a slip as shown by. the context. Mr. Pryor makes the follow- 

 ing statement about^. aqnatica which at first seems unintelligible: 

 " The anthers may be seen ijx) projecting some distance above the 

 lower lip of the flower, while the style is still undeveloped, and con- 

 cealed some way down in the tube. In other cases (J)) the stamens 

 and styles alike have risen, but not to the same extent as in a, above 

 the level of the surface; while in a third instance (r) the anthers are 

 barely visible, but the greatly lengthened style is closely reflcxed 

 over the lower lip of the corolla, and indeed almost reaches the 

 upper border of the sepals. Are these several changes only the 

 result of growth, oris there anything like trimorphism in'these plants.? 

 More probably, perhaps, as in case of some of their nearest allies, 

 they are regularly protandrous, and these appearances would thus 

 find a ready explanation" (i, p. 259). Stages b and c are readily under- 



