Botanical Affinities t/ Orobanche. 41 1 



as a mere residuum of the nutrient mucilage in which the em- 

 bryo is originally generated. If this be so, then it will follow 

 that the excessive abundance of albumen is a high physiolo- 

 gical character, which must be considered paramount to all 

 peculiarities in the mere arrangement of the floral organs with 

 respect to each other; lor the latter, so far at least as the 

 calyx and corolla are concerned, cannot be supposed to have 

 any essential connection with the function of reproduction. 



Now Orobanchacea? differ from Scrophulariacea? in their em- 

 bryo being very minute and their albumen extremely copious. 

 In these respects they again correspond with Gentianacea?. 



Altogether agreeing with Orobanchacea? in the large mass 

 of their albumen, and in their brown scaly parasitical habit, 

 we find Monotropacea?, placed in a distant part of the system 

 by everybody except Horaninow. To this botanists have 

 been led by the hypogynous stamens of the one order and the 

 epipetalous stamens or the other; a character which so often 

 produces artificial results, and which, I think, even the French 

 school will be compelled to abandon as a great means of 

 systematic distinction. 



The supposed relationship ofMonotropa to Ericacea? through 

 Pyrola, may seem an objection to the approximation of the 

 former genus to Orobanche and Gentianacea? \ and if it were 

 certain that Pyrola was related to Ericacea the objection 

 would be a good one. But Pyrola differs from Erica cea? in 

 its albumen, just as Gentianacea? and their allied orders differ 

 from Scrophulariacea? \ and I cannot but regard the placing 

 Ericacea? and Pyrolacea? in contact as another of those false 

 approximations by which our ideas of classification are so 

 much deranged. This is not the place to enter upon such a 

 discussion, or I should be prepared to show that Pyrolacea? 

 are more nearly allied to Droseracea?, and Ericacea? to Rntacea?, 

 than to each other. Pyrolacea? are related to Ericacea? chiefly 

 by their hypogynous stamens, porous spurred anthers and 

 indefinite seeds, for it is an arbitrary application of terms to 

 call them monopetalous. They are different in their whole 

 habit, the organization of their seeds is essentially dissimilar, 

 and the Monotropacea? have not even porous anthers. But 

 on the other hand they stand nearly related to Orobanchacea? 

 in their tendency to become leafless and parasitical, in the 

 placentae being half parietal in Monotropa, in their anthers 

 having usually the spurs of Orobanche, and most especially in 

 the proportion borne by the albumen to the embryo. 



I would therefore submit that, for the reasons now assigned, 

 Pyrolacea? (including Monotropacea?), Orobanchacea? and Gen- 

 tianacea?, instead of standing distantly apart in the natural 



3 G 2 



