Baron Humboldt 07 i Roclc Formations. 
cumulation of carbon or nodules of compact limestone in the 
clay-slates, nigrine and epidote in the syenites, (alternating with 
granite and porphyries), conglomerates or pudding-stones, ha- 
ving a basis of anthracitic mica-slate, may, without doubt, be 
characteristic of transition formations ; in the same way as, ac- 
cording to the useful labours of M. Brongniart, petrifactions of 
shells, in a good state of preservation, sometimes indicate direct- 
ly such or such a bed of tertiary deposits. But these cases, 
where we are guided by disseminated substances or by charac- 
ters purely geological, comprehend but a small number of 
rocks of a recent origin, and observations of this kind often lead 
only to negative facts. The characters taken from the colour, 
from the grain, and from small veins of carbonate of lime, 
which traverse calcareous rocks ; those which are furnished by 
the fissility and silky lustre of clay-slate, the aspect and undu- 
lations more or less marked of the scales of mica in mica-slate ; 
and, lastly, the size and colouring of the crystals of felspar in 
the granites of dilferent formations, may, like all that is con- 
nected simply with the iihysiognomy of minerals, lead the most 
expert observer into error. The white and black tints un- 
doubtedly in most instances distinguish the primitive and tran- 
sition limestones; the Jura formation, especially in its upper 
beds, is also without doubt generally divided into thin whitish 
beds, having a dull, even, or conchoidal fracture, with very 
flat cavities : but in the mountains of transition limestone there 
are isolated masses which, in colour and texture, resemble the 
oryctognostic characters of the Jura limestone ; and to the south 
of the Alps there are hills belonging to tertiary deposits, where 
we find rocks analogous to the slaty and dull Jura limestone 
(in as far as regards appearance,) in formations placed above the 
chalk, and which resemble the limestone used for lithographic 
purposes. Were names taken from their oryctognostical cha- 
racters alone to be preferred in distinguishing formations, the 
different strata of the same compound rock having a consider- 
able thickness, and extended to a great length in a particular 
dhection, would often seem to belong to diflerent rocks, accord- 
ing to the points at which specimens were taken. Consequent- 
ly we can only determine geognostically in collections, suites 
rocks of which the mutual superposition is known. 
