w 



ANALOGIES AND AFFINITIES. 45 



there almost prevailingly numerous. Very mauy are the 

 species now known— and they distributed through several 



genera— whose six stamens are of precisely equal length; 



while in a number of capparids the tetradynamy, formerly 

 tliought of as limited to ciucifers, is not in the least obscure. 

 In Vclla psciido Cytisiis, a crucifer, it is perhaps only 

 emphasized by the circumstance that the four long stamens 

 are in two monadelphous pairs ; but in a lengthy series of 

 Pacific North American species which have been placed in 

 different genera all far too near Arahis, the stamens are in 

 three very unequal pairs, the longest pair monadelphous 

 sometimes near the base of the filaments only, but often 

 almost up to the base of the anthers. The numerical analogy 

 of tetradynamy is, indeed, preserved ; the character itself is 

 gone. But in M€gac^rj:oca polyandra, unquestionably a 

 crucifer, the stamens are more than six— even indefinitely 

 numerous. However, it is in the distinctive kind of hypogy- 

 nous insertion of the stamens in the family, that we find the 

 n-arest approach to an absolute floral character ; but this 

 fails signally in Suhularia, the stamens of which are not 

 hypogynous at all, but clearly perigynous, the sepals cohering 



to form a calyx-tube. 



The regularly cruciform spreading of the petals, a circum- 

 stance long ago suggestive of the name which the order now 

 bears, may not have been thought of at any period in the 

 history of the science, as a fundamental characteristic. 

 Irregularity in this floral circle in Ihcris and other Old 

 World genera was always known. But I may here take occa- 

 sion to speak of a curious way in which the cruciform aspect 

 of the corolla is obscured throughout a lengthy ^series of 

 Pacific North American species ; a striking pecmianty ot 

 them as it appears in the living plants. In all of them- 

 Slrcpfanfhns, Canlanfhu.., some of the kinds of Thdupodnim 

 etc.-the blades of the petals are narrowed extremely and set 

 apart, into two pairs, an upper and a lower so that the 

 corolla as a whole is in a manner bilabiate, the upper lip 

 and the lower, so to speak, being rather widely separated, 



