620 ON THE DISTINCTION OF SPECIES. 
fact is, that the young plants of B. alba bear completely cor- 
date leaves: the leading shoots of older trees, as well as the 
fast-growing shoots which come from lopped branches or 
the lower part of old stems, also produce cordate leaves; 
while the weaker, and often drooping twigs on the very same 
trees, are clothed with rhomboid or triangular leaves. More- 
over, the leaves and bark of the young plants are usually, if 
not always, pubescent, as also is the case with those of strong 
shoots on grown trees. These facts lead to the presumption 
of Betula alba, glutinosa, and pubescens being only forms of _ 
a single natural species. Of the alleged difference in the 
fruit and serratures of the leaves, I will state something here- 
after. 
|... Fifthly the remarks on Betula alba naturally lead to 
another rule ; namely, that contrasts of characters should be 
made between parts which are at equal stages of growth, and 
under equal conditions of luxuriance : if made otherwise, 
they are likely to prove fallacious. Thus, it is stated that 
the style of Erica Mackaii is “ protruded ;” that of Erica Te- 
iraliv being “usually included." In E. Tetralix (and pro- 
bably in E. Mackaii) the corolla and pistillum are equal when 
the flower first opens, the pistillum gradually elongating 
afterwards until the corolla fades, when the styleis protruded 
by about a fourth of its length. The included or protruded 
style is thus a condition of age, and valueless as a distinctive 
character. As an example of the propriety of making 
contrasts only between plants or parts in equal conditions of 
luxuriance, I wil take Cerastium alpinum and Cerastium 
latifolium, of the British Flora. In the Manual, the latter 
is reduced (incorrectly, I think) to a variety of C. alpinum; 
but they are both shortly described there, and in their de- 
scriptions it is stated that the stems of C. alpinum are 
* mostly simple," those of C. latifolium “branched?’ On 
the contrary, it is said in the British Flora, that C. alpinum 
is * much branched." The fact is, both species are much 
branched ; C. alpinum, perhaps the most so: and when grow- 
