ON THE TWO VALERIANS. 848 



On June 11th the tallest stem of Mikanii measured 4ft. 2 in., 

 that of sambucifolia 4 ft. 5 in. ; ultimately two or three stems of 

 Mikanii exceeded the tallest of sambucifolia by a few inches. The 

 stem of Mikanii is much (probably one- third) more slender than 

 that of the other. The colour of the foliage, stems, &c, was very 

 different throughout the growth of the two plants. In Mikanii the 

 leaves were of a dark, opaque, somewhat bluish green ; in sambuci- 

 folia of a beautiful, bright, clear green, much like that of Cardamine 

 amara. The more divaricate habit of Mikanii, already mentioned 

 when speaking of the lower leaves, was characteristic throughout 

 its growth ; it is seen in the more widely- spreading main branches 

 of the inflorescence, and reaches a climax in the fruiting cymes. 

 In the ultimate cymes the branchlets are more than divaricate 

 they may almost be called scorpioid ; in sambucifolia the ultimate 

 branchlets remain simply erect. There is an appreciable difference 

 in the shape of the ripe fruit; that of sambucifolia is actually 

 larger, considerably broader at the base in proportion to the apex, 

 and the seed does not fill the cavity, so that, the empty margins of 

 the fruit being appressed, it has a kind of false wing on each side 

 in the lower part. The fruit of Mikanii is smaller, more oblong in 

 form, less tapering upwards, so that there is less difference in its 

 breadth at base and apex ; this is partly owing to absence of the 

 wing-like expansion of the pericarp, which is only sufficiently wide 

 to enclose the seed which entirely fills the cavity. (No fruits 

 should be compared which do not contain ripe seed ; sometimes 

 many are sterile.) Strictly speaking, V. sambucifolia. is normally 

 a soboliferous plant, and I have never seen it with aerial stolons ; 

 once in a dried-up swamp, where the ground was very hard and 

 caked, the soboles crept along the surface, but half underground, as 

 though they would have been quite underground, if possible. 

 V. Mikanii seems more variable ; it is often soboliferous, but some- 

 times, on the chalk, produces stolons which are not only aerial, but 

 eomewhat arching, so that they only touch the soil again at the 

 footing apex. In cultivation both plants were at first soboliferous 

 in habit, and sambucifolia remained so ; but in Mikanii, after some 

 growth had been made at the end of the soboles, the intermediate 

 Part rose above ground and became strongly arched, and often con- 

 torted in fantastic shapes. This, and the arched stolons sometimes 

 seen on the chalk, seem to point to some different inclination in the 

 two plants, but I am unable to frame any character therefrom. 

 The leaflets of sambucifolia are variable, and there is a sub-entire 

 leaved form ; in Mikanii the autumnal states and the young plants 

 on barren stolons, sometimes produce leaves which are a good deal 

 more cut than is normal ; I think there is less variation in the 

 number of the leaflets than in their cutting. In sambucifolia 1 tad 

 the well-developed leaves to have commonly 4-6 pairs of leaflets, 

 and I cannot discover, in my large collection of forms of this plant 

 from both wet and dry places, a single example in which there are 

 more than six pairs ; in Mikanii they are commonly 8-10, though 

 8 onietimes fewer in ill-developed leaves ; the root-leaves of both 



