DR. M. T. MASTEKS OS THE MOKPIIOLOGY OF THE MALVALES. 25 



structure in question by considering the petals to constitute a 

 portion of tlie androeciuni, and not a true corolla, just in the same 

 way as the petals of Sosacece are considered by Dr. Alexander 

 Dickson to beloug to the staminal whorl*; and this notion is in- 

 directly supported by the frequent absence of the corolla in 

 Sfercniia, or its minuteness^ as in GiiicJienotia &c. But this 

 view is not borne out by organogeny, it being pretty clearly 

 ascertained that the petals and stamens are distinct in their 

 origin, and that the former precede the latter in their develop- 

 ment. This original separation of the two whorls also precludes 

 the acceptance of a third hypothesis, that the stamens are pro- 

 duced from the petals by chorisis or clcdouhlement. 



The hypothesis which appears to me to have the greatest amount 

 of support from comparative morphology or analogy, is one accord- 

 ing to which the key to the whole structure would be afforded 

 by the presence and relative position of the staminodes. These 

 are generally five in number, and almost invariably opposite to 

 the sepals, thus occupying the normal position of the stamens. 

 Payer, indeed, states that, in Pavonia^ the sterile teeth of the 

 column are antipetalous, an assertion of the truth of which 

 I have not been able hitherto to convince myself, though in some 

 Tiliacecd the staminodes are superposed to the petals. But even if 

 it be so, the probable existence of more than one row of stamens in 

 this genus would still leave my hypothesis unaffected. I consider 

 that the central lobe of the staminal leaf is placed almost inva- 

 riably, if not always opposite to a sepal, that sometimes this c 

 tral lobe is fertile, /, e, antheriferous, and then it does not difier 

 from the ordinary stamens ; at other times it is sterile and peta- 

 loid, always traversed by a central rib, and then receives the name 

 of staminode. Such a case as WaUheria, wliere there is a single 

 stamen in front of a single petal, may be explained by the j^artial 

 suppression of the central and of one of the lateral lobes of the 

 stamen, one lateral filament alone being developed. Traces of 



the suppressed organs may, indeed, be seen in young flower-buds 

 at the very base of the flower ; and it is probable that they would 

 be more clearly seen in fresh flowers ; at any rate it is easy 

 to see that the stamen is not exactly central. Moreover, al- 

 though, as a general rule, where the staminode is immediately in 

 front of a sepal and has on each side of it one, two, or more sta- 

 mens it may be considered to represent the terminal or central 

 lobe of a digitate leaf, yet in other cases it may be that the 



* Sccmaitn's Journal of Botany, toI. iii. p. 216, 



^71 — 



