284 Scientific Intelligence. 



above referred to — among other matters of high consideration, Mr. Ben- 

 tham works out his hypothesis of glandular stamens with some detail, in 

 explanation of the androeoium of certain Sterculiacece. We are not pre- 

 pHred to discuss the points here involved, not only because this would 

 require time and space which we cannot now afford, but also more knowl- 

 edge than we can now pretend to. But we would remark:— 



(1.) That we are well satisfied in finding that Mr. Bentham fully ac- 

 cepts the view that a single leaf of the andrcecium may be represented by 

 a phalanx of stamens (this, indeed, he had previously favored), or by » 

 cluster of stamens and a scale or petaloid body, such for instance as we 

 have m an American Tilia ; also " that a dedoublement of the petal may 

 produce the inner petaloid scale of some Sapindacea>., Violacece, Bixacea, 

 &c., or the fimbriate scales in the tube of Cuscuta and other gamopelal- 

 ous flowers." That is about as far as we incline to go in this direction. 

 But, as Mr. Bentham homologizes these scales or appendages with petic- 



greatly to object when some convert to his theory imagines sutib poten- 

 tial anthers to become actual ones. 



(2.) But Mr. Bentham's ingenious hypothesis to account for the ante- 

 position of the outermost stamens to the petals in Sterculiacea, &c., »» 

 deduced from Glossoslemon, apparently would work just as well on the 

 supposition that the anthers answer to the leaflets of a palraately com- 



(3.) Sidalcea, that most instructive Malvaceous genus, and a pentandrous 



other, taken together, offer to our view most serious if not conclusive 

 objections to the conjecture that, in the family in question, the real or the 



Nor do we see how this view applies "to Melochia and its allies, where 

 only a smgle stamen stands before each petal, unless these stamens are 



(4.) Apropos to the latter, and much inclined to accept Mr. Benthain's 

 views in regard to the androeciumof Bombacea, we remark that the p''" 



Cheirostemon, and Uriodendron, is before the sepals, instead of being be- 

 fore the petals as may be deduced to be the case in the true i/aVfOC^ 

 no less than in Sterculiacecs and in Tilia. Whatever hypothesis W 

 axlopted in explanation, the position of the stamens relative to adjacent 

 parts of the flower is important. On this and other accounts, in the an- 

 nexation of the Bombaceas to Malvacece, we could not hesitate to accord 

 to the group the rank of a suborder. -*- '^■ 



^^ 2. Wood-eells of Hamamelidem imitate Coniferous markings.— ^^ P''P^' 

 On Sycopsisr a new Hamamelideous genus, published by P'^f^^'^^; 

 J>l»ver, in the 23d volume of the Linnsean Transactions, brings to vie^" •^ 

 lact discovered by the late Mr. Griffith, in Bucklandia, and ann^"!""^ 

 tor ai the other genera of this order, that the wood-cells exhibit ma*ng» 

 inl' /^P-^/^^ ^"^^^ ''™''*'" to those of the so-called 'glandular m»* 

 togs of coniferous wood." As Prof. Oliver bints at an affinity to or po^ 



