In both cases filmy white mica is found on the cleavage-planes, and the phenomenon is designated 'cleavage-foliation.'

Metataxis is recorded in the wavy contortions of the original foliation-planes and the approximation of the flakes of mica to a parallelism with the cleavageplanes.

Figs. 3 and 4 also furnish excellent examples of subsequent metataxic alteration.

With these we may compare the statement of Allport (Q.J.G.S., vol. xxxii, p. 426) that "there is evidence to show that the slates of the Cornish Peninsula existed as metamorphic rocks (i.e. cleaved and in places contorted) long before the intrusion of granite. There the contact metamorphism extends to a short distance only (quite distinguishable by modern methods of research from the previous metamorphism) the transition from the one to the other being very gradual."

The importance of the unconformity referred to on p. 17 has been emphasized by Prestwich ('Geology,' vol. i, p. 419).

(1) The passage 'by insensible gradations' of the 'early gneisses' on the one hand into granite and on the other into mica-schists, with which crystalline limestones quartzites and iron-ores are instratified, is contrasted with

(2) the fact that "between the Archæan rocks and the succeeding Cambrian series there is in Europe as well as in America a marked break of continuity."

How disappointing it is in connection with this subject to turn to even the more recent of our text-books written by our foremost geologists, may be seen by noting that in Geikie's Text-book of Geology the archæan crystalline schists are dismissed within a page and a half (pp. 588, 9). The foregoing facts and arguments appear to lie altogether outside the mental horizon of the distinguished author of that work. It is entirely to miss the point to talk of these rocks having been formed during a period of the earth's history when the ocean had a considerably different relative proportion of mineral substances dissolved in its waters "; or, again, to speak of "the same order (of chemical precipitation) having been followed everywhere over the floor of the ocean"; and that for the simple reason that the theory requires, and deduces from known facts and principles, a prevalent set of physical conditions under which the ocean could not possibly have existed as such * The vast range of temperature between the first initial oxidation of hydrogen † and the condensation of steam into water on any general scale must have afforded ample space in time in the cooling-down of the original nebulous mass of the globe for the formation and deposition of silicates and the separation-out of the quartz, which together make up the granitoid rocks, the gneisses, and the crystalline schists; while on the other hand the high degree of stability of most of the silicates removes all difficulty in the way of their formation while, as yet, the temperature was too high for the condensation of water in anything like oceanic

* cf. Pfaff, op. cit. pp. 25 et seq. See Appendix ii. Note A.

proportions. These considerations lead us to contemplate rather a condition of things in which deposition on a vast scale was going on from the then dense atmosphere, so as to form a growing non-consolidated lithosphere on the surface of the globe-an universal hot magma, which by gradually crystallizing would leave traces of a sort of bedding arrangement of its materials from differences of the specific gravity of the minerals formed,* while the mechanical process of stratification in aqueous basins (as the term is generally applied) would be out of the question. Such a view, when allowance is made for contemporaneous paramorphic change by chemical reactions between the minerals, appears to be a partial explanation of the banded structure of the archæan schists, when viewed on a large scale, and the occurrence in them of quartz-layers often of considerable thickness.

A similar incapacity for comprehending the theory here advocated appears in Green's Physical Geology' (pp. 426-27.) Nothing is easier than to dismiss as 'dreamy conjecture' what we do not comprehend; and Mr. Green's misconception of the theory is shown by his attempt to state it in the following words: "The constituents of those rocks were supposed to have been held in solution in an ocean of boiling water, and to have been precipitated as it cooled."t

I submit that the argument advanced in this section of the present work proceeds by way of induction from known facts; and that the further fact that there is a tendency in some quarters to depreciate an argument which proceeds on chemical and physical lines is no refutation of the theory advanced.

Deformation of rocks by the shearing and crushing which have inevitably accompanied the great earth-movements, whose effect is traced in enormous foldings, fractures and overfolds, (by the resolution locally of normal into tangential forces in the contraction of the Earth's mass in the process of cooling) has certainly brought about petrological changes; but they do not appear to have resulted anywhere in the formation of a true schist or gneiss out of originally clastic rocks. This point is more fully discussed later on. Cotta pointed out in some remarks on the contorted strata about Vierwaldstädtersee (in a letter dated from Zürich in 1849) that the strata which we now see contorted must have been at the time in a more pasty condition from saturation with water, and probably have acquired their present rigidity by partial desiccation since their upheaval. Yet even in this pasty condition pressure has not succeeded in converting them into crystalline schists. No rocks on a large scale

* See Section v.

+ The absurdity of this notion is exposed by Pfaff, (op. cit. p. 27.)

are equally saturated with water; and in great earth-movements it follows that with different degrees of plasticity (owing partly to this fact and partly to differences of constitution) it would necessarily result, that, while some portions were bent, contorted, and folded, other portions would suffer fracture, crushing, and in places even pulverization in different degrees. In such portions we should be prepared to find clastic materials present; but we have clearly no right whatever to infer from this fact the derivation of a deformed crystalline rock-mass from originally clastic materials. One of the main props of the theory of extreme regional metamorphism is thus knocked away.*

The simulation of the macroscopic characters of true schists by later rocks as the result, partly on the one hand of hydrothermal and chemical action, partly of deformation of rocks, by pressure, crushing, and shearing on the other, leaves open still a vast field for further microscopic research, united with field work.t

Among the accessory minerals common in the oldest crystalline rocks the most widely distributed are apatite, rutile, zirkon. The occurrence of these very stable minerals as accessories among the primal mineral constituents of the oldest rocks (granite, syenite, gneiss, mica-schists) as well as in the chief varieties of eruptive rocks, is entirely in accord with, and thus far lends support to, the theory that relative stability has been the main factor in determining (of course in the inverse order) the general succession of mineral deposits in the formation of the earth's crust.

It is not contended that in this there has been anything like absolute uniformity of succession. That we should consider on a priori grounds extremely unlikely, because we have no grounds whatever for assuming the thermal energy which pervaded the non-differentiated mass to have been equally distributed throughout it.‡ And so it happens-as theory would lead us to expect-that numerous irregularities are observed in the general order of succession; now this now that accessory mineral being common in the various types of the fundamental rocks; schists occurring subordinately among the gneisses, and gneiss occurring subordinately among the schists; garnet taking the place of mica, and giving us

* Since this was written a remarkable instance of the effect of crushing and subsequent decomposition of the pulverised materials in producing a pseudo-breccia in the gneissic rocks of the Malvern Hills has been described by Professor McKenny Hughes. (Geological Magazine, November, 1887).

Kalkowsky (Lithologie, p. 43) remarks: "In Folge von Pressungen und Spannungen bei der Bildung können Körper dieser Klasse (einfach brechende Körper) doppelt licht-brechend sein, meist aber nur in schwachem Grade. Es bedarf eines sehr guten Polarizationsapparates und eines unermüdeten Auges, um solche Spuren von Doppelbrechung wahrzunehmen."

See App. ii. Note R.

locally granulite in the place of granite; gneiss containing in different regions, some here others there, such accessories as graphite, garnet, tourmaline, epidote, rutile, zirkon, hornblende, chlorite, apatite, micaceous iron-oxide, magnetic iron oxide, pyrites; the mica contained in gneiss though generally muscovite, being sometimes biotite (Erzgebirge), as is the case occasionally with granite (Dartmoor); hornblende in one variety of gneiss, cordierite in another, augite in a third, garnet, graphite,*~ chlorite, green mica, all in their several turns, replacing, or appearing side by side with, the mica, and giving their respective varietal characters to the gneisses.

Yet through all this apparent confusion there is traceable a leading principle in the oldest rocks—general predominance of the more acid minerals (the most stable) in the granitoid and gneissic rocks, while, as we pass into the overlying schists and phyllites of the archæan series, the more acid minerals gradually give place more and more to the more basic minerals which are somewhat inferior (as a class) in stability to the minerals which preponderate in the granites and gneisses.

We have been here concerned (under the head of 'paramorphism') with a number of mineral changes, of which chemistry gives us some knowledge, among the materials which make up the mass of a rock per se; and we have not limited our view to changes which might occur after the rock was once formed, but have rather endeavoured-in a firm belief in the continuity of nature's principles of operation—to trace back the working of the principles learnt from physics, chemistry and mineralogy to the primal genesis of the most ancient crystalline rocks,— the granitoid, gneissic, and schistoid rocks of the earlier crust of the globe. In this we have arrived independently at some general conclusions which are found to be to some degree in accord with Gümbel's doctrine of 'diagenesis.' In all such changes we may regard the mass of the rock as remaining pretty constant in its composition after the primal accumulations of its materials, notwithstanding the changes which may take place reciprocally among its mineral constituents; so that a series of bulk-analyses would—could they have been made at the various stages of its history-have given pretty constant results. It is in this sense that one feels justified in the use of the term "paramorphism" adopted here.

Primary paramorphism has thus been treated as having to do mainly with the actual genesis of the rocks themselves through the agency of mineral change; and in this sense it comes within the larger subject of "metamorphism." We have been led to see strong reasons for regarding the processes

* See Appendix ii. Note L.

Kohlensäure, die heute den Karbonatgesteinen als solche, in den Pflanzen und Kohlengesteinen, sowie im Graphit als Kohlenstoff, im Bitumen als Kohlenwasserstoff der Erde einverleibt ist, noch in gasförmigem Zustande verteilt befand. Diese an Kohlensäure reiche Regenwasser mögen auf die Kalk-und Magnesiasilikate der Erstarrungskruste in hohem Grade zerlegend eingewirkt und dem Meere stark konzentrirte Solutionen von Kalk- und Magnesiakarbonaten zugeführt haben."

In Note L (App. ii) I have shown how our laboratory-experience leads us to see the probability of the production of the archæan graphite by dissociation of hydrocarbons; but with this exception we may accept the idea contained in this passage in its entirety. The same writer points out how the absence of organisms capable of secreting carbonate of lime from these primeval waters would be favourable to their local concentration; he also notes that the water of the present ocean requires evaporation to the extent of 75% before precipitation of CaCO3 sets in.

Accepting Bischof's statement "that the carbonate of lime of all the formations would form a layer over the Earth 1000 feet thick," Pfaff (Geol. als Ex. Wiss., p. 162) has calculated that the CO2 contained in it alone would in the free state give a pressure of 356 atmospheres at the surface of the Earth. This is probably excessive if in the calculation the force of gravity has been assumed the same as we know it.

The action of CO2 appears to me to have reached its maximum in later archæan and earlier palaeozoic times, the grauwacke of the former forming perhaps a connecting physical link (with its authigenous mica) between the archæan crystalline rocks and the true sedimentary palæozoic series. This, I find, has been recently pointed out by Kalkowsky (Elemente der Lithologie, p. 280.) "Dass diese Gesteine (archäische Grauwacke, including 'Glimmertrapp') eine Brücke bilden, die uns zu den holokrystallinen archäischen Gesteinen hinüberführt, so dass über diese neues Licht verbreitet wird, ist augenscheinlich."

I have adopted the terms 'authigenous' and 'allothigenous' from Kalkowsky (Lithologie, p. 13) because they seem exactly the words wanted. 'Endogenous' and 'exogenous' are terms that appear to have found favour in some quarters in this country; but the two serious objections to their use in this sense are, (1) that the connotation they bring with them from morphological botany is misleading: (2) they have been appropriated in a different sense in lithology (see note to § vi).

May we not account for the occurrence of the authigenous micas and the garnets in the cement of these grauwacke as resulting from the combination of the nascent SiO2 (replaced by free CO2 as suggested in this work p. 13) with the requisite bases sparingly present, while the large excess of the SiO2 went to form the bulk of the grauwacke?

Such considerations point to a very unequal distribution of the silicates in the earlier crust, as well as marked localization of the earliest condensed waters; but the action of free CO2 may be regarded as perhaps the main factor in determining the distribution of the earliest limestones and quartzites. Where the CO2 attacked MgSiOs or CaSiOs (or both) impure (siliceous) carbonates would be formed; but where Na2SiO3 or K2SiO3 were chiefly present the carbonates formed would be removed in solution, and the SiO2 would furnish the mass of the material of a grauwacke.

The distribution of the limestones and quartzites (if such was the mode of their genesis) shows the continued action of free CO2 as a very potent paramorphic agency in earlier paleozoic times; while the theory dispenses entirely with all necessity for postulating the existence of living organisms as agents concerned in their production.

The fact that excess of CO2 in solution in water causes CaCO to enter into the form of the soluble bicarbonate CaH2 (CO3)2 may appear prima facie to be a fatal objection to the theory that limestones per se could be formed in any quantity by the direct action of carbonic acid on the lime- and magnesia