Germination and Growth of Species of Salvia. 367 



early date and those of the more mature plant. Eegarding 

 changes in shape, perhaps no general statement can be made, 

 except that it is at times a change from a simple leaf to 

 a compound one — that a simple leaf always precedes a lobed 

 or finely-divided one. This is seen not ouly in the young 

 plants as they develop from seed, but it is shown in almost 

 every mature plant on the different axes, where the simplest 

 forms of leaves are those produced earliest ; that is, those 

 outermost in the bud. 



Examples come to mind of the change that may be 

 noticed in countless Leguminosse, in Scahiosa, in Cineraria 

 maritima, and others. 



Perhaps, more distinct than such alteration in shape is the 

 very evident alteration in the nature of the surface, especially 

 in those species which are of appressed habit. This change, 

 in general, is one from an almost characterless to a special- 

 ised leaf, from one without more than a light-coloured 

 depression for midrib, with a few large veins given off on 

 each side, to a leaf which is rugose, with a closely-netted 

 venation and coarse veins on its under surface, and a wavy, 

 creuate, or serrate leaf margin. 



The features of most marked description of which we 

 must now take notice are those which give to the plant its 

 general habit ; by this we shall roughly classify our species, 

 calling attention to certain correlated or, at least, accompany- 

 ing characters. 



The roots of the species we have examined are all very 

 much alike, in most species the primary root thickening very 

 evidently, and from this small, wiry, secondary roots are 

 given off in the upper part, and longer, somewhat more 

 vigorous fibrous roots from lower down. In some species, as 

 S. Forskohlei, the upper part of the primary root is very much 

 thickened, giving the appearance of a tap root at first ; 

 in general it is wiry and branched at a distance of sixty 

 to eighty mm. from the " crown." 



S. canariensis apparently formed much less vigorous roots 

 in comparison with the size of its leaves than any other 

 species ; the secondary roots were more wiry than was 

 observed in other species. aS'. lyrata and S. ohovata possessed 

 roots in which the primary root was not so evidently 

 thickened, others of almost equal size arising from its upper 



