THE SEEDLTNO FOLIAGE OF ULEX OALLIT 11 



pairs trlfoliolate and bifoliolate leaves occurred in about the same 

 numbers. The more developed pair-types 3 + 8, 8 + 2, 8-1-1, and 

 2 + 2 diminished progressively from the first pair to the sixth (with 

 a slight reversal in the fourth pair in the case of 8 + 2 and 8+1), 

 The less-developed tj^pe 2 + 1 attained a maximum in the second pair 

 and duninished onwards (with a slight reversal in the fifth pair). 



It does not appear from the accounts given by Lubl)ock and 

 Boodle of the seedlings of TI. europcBus whether the trifoliolate con- 

 dition is any commoner in the first pair than in the ones immediately 

 following. In U. Gallii, however, the trifoliolate condition is 

 realized in the first pair in nearly 75 per cent, of all cases, and pro- 

 gressively less in the subsequent pairs, all trace of compound leaves 

 disappearing after the sixth pair. 



The sub-tribe Cijtisince includes the four genera Kypocalijpiv^, 

 Loddigesia, Gytisus, and TJlex. The two former have trifoliolate 

 leaves, some species of Cytisus have trifoliolate and others have 

 simple leaves, and TJlex normally bears only simple leaves on the 

 adult plant. Few botanists w^ill be inclined to dispute that the 

 ancestral leaf-condition of the Ctjtisinas was trifoliolate, unless they 

 accept the view that compound leaves are derived from simple ones 

 (G. Henslow, Orig. PI. Struct. 246; 1895). Some idea of the 

 probable course of leaf -reduction in TJlex may be gained by com- 

 paring the foliage of Cytisus scoparhis with that of IT. euroj^ceus 

 and IT. Gallii. C. scoimrkis usually has trifoliolate leaves on the 

 main stem and branches, and simple ones on the final branchlets, the 

 extent to which trifoliolate leaves develop apparently depending to 

 some extent on the environment. Bifoliolate leaves often occur 

 between the trifoliolate and simple ones. In the next stage of reduc- 

 tion the compound leaves may have been confined to the low^r part of 

 the main stem. In TI. europtsus they are usually restricted to the 

 seedling foliage, namely to the leaves between the cotyledons and 

 the first spiniform leaves. Finally, in U. Gallii they are largely 

 confined to the first two pairs of leaves following the cotyledons, and 

 it is only in the first pair that they are more frequent than simple 

 leaves. 



Note. — After the foregoing was in type, my attention was called 

 by Mr. L. A. Boodle to a paper by H. Wager, " Observations on the 

 Morphology of Species of the Genus TJlex'' {Inferncdional Journal 

 of Microscopy and Natural Science, January 1897). This was not 

 cited by Mr. Boodle (Ann. Bot. xxviii. 527; 1914), as he was 

 unaware of its existence at the time of writing his paper. Wager 

 " found that taking a large number of seedlings [of TJ. eiiropmis'] 

 from two equally exposed but different soils, one humus and the other 

 sandy loam, the percentage of seedlings with trifoliolate leaves is not 

 only "greater on humus soil than on the sandy loam, but the spinescent 

 character is more quickly assumed in the latter case than in the 

 former " (reprint, p. 9). " In a normal seedling the cotyledons are 

 succeeded by one or two pairs of trifoliate leaves .... succeeded by 

 several pairs of spathulate leaves .... These first leaves may be 



