Goldschmidt’s “Grundlagen der quantitativen Geochemie” (1933)

 Introduction

Geochemistry, a branch of geological sciences, started its existence with contributions of Frank Clarke, Victor Goldschmidt and Vladimir Vernadsky. The initial concern of geochemistry was to characterize the chemical composition of the planet. Nowadays papers by Condie (1993), McDonough and Sun (1995), Taylor and McLennan (1985), Ronov and Yaroshevsky (1969), Rudnick and Fountain (1995), Turekian and Wedepohl (1961), Vinogradov (1962) and others are probably the most relevant references in the field. However, early influential contributions regarding the Earth’s composition were papers by Clarke and Washington (1924), and Goldsmith (1933). Their studies were focused the average composition of the most accessible part of the planet – continental crust and its components (e.g., igneous rocks merely). While, the paper by Clark and Washington (1924) is easily accessible on the internet, well-cited Goldsmith’s “Grundlagen der quantitativen Geochemie” (1933) was accessible for me only via library order. The actual reference is: Goldschmidt, V.M. (1933). Grundlagen der quantitativen Geochemie. Fortschrift Mineralogie 17(2), 112-156. Of course, it comes in German. I wondered how often such an influential paper was actually read (although I am aware of later Goldschmidt’s papers and a book in English, they are rarely cited in the context).

A good friend of mine, Sara Yanny-Tillar, recently received her Master’s degree in Germanic languages from the University of Illinois, Urbana-Champaign. Her interest in German language is admirable and she was very kind to help me with translating a part of the paper. The translation turned out to be great and Sara said it was good experience for her to translate something scientific. Thank you, Sara!

Translating the chapter “Durchschnittliche Zusammensetzung der Eruptivgesteine” (Average Composition of Igneous Rocks) is especially important as Goldschmidt used the new approach to estimate the average composition of igneous rocks and continental crust overall. Arguing that the method used in Clarke and Washington misinterpreted the proportions of rock composing the average continental crust, Goldschmidt used fine-grained sedimentary rocks such as post-glacial tillites and shales as they naturally preserve proportions of rocks composing the crust.

 

Fundamentals of Quantitative Geochemistry, by V.M. Goldschmidt (Göttingen)

IV. Average Composition of Volcanic Rock

The figures of Clarke and Washington are primarily cited for the average composition of volcanic rock (as the average of all prior conducted analyses on recent volcanic rock). Corresponding average values have also been calculated for the different continents and individual countries. A fundamental objection has been raised against the method of calculation applied by Clarke and Washington, namely, that the calculation of an average from collected analyses of volcanic rock, in which no attention was given to the proportions of the rocks, naturally produces a statistical bias towards rare and petrographically unusual types of rock, in contrast to more common, and more extensively dispersed types of rocks. For example, rare varieties of nepheline-bearing rocks, which often only appear in very small amounts, are given too much attention in the calculations compared to the significantly large amounts of the more common granite. This bias can also be observed in Washington’s calculation of the average composition of Norwegian rocks, in which the overwhelming portion of analysis comes from the relatively small amounts of alkaline rocks in the Oslo region, while the large amounts of Precambrian granite and quartz diorite are only represented in analyses in a limited scope. J.J. Sederholm (1925) most notably pointed out this imperfection of the method of calculation by Clarke and Washington.

We can, however, check the usability of the average values from Clarke and Washington by examining the sedimentary material, that, after its formation, provides a useful average value of the crystalline rock of a larger region of the earth’s surface area. The most applicable is the material of the glacial and post-glacial clay deposits in southern Norway: an area of about 200,000 km2 that is overwhelmingly comprised of crystalline rock. A primarily mechanical comminution took place during the ice-age era erosion of the region; chemical weathering occurred only on a relatively low-scale. The glacial and post-glacial clay deposits are consequently comprised to a large extent of only mechanically comminuted rock material. A considerable number of complete analyses have been carried out on large average samples of this clay in the mineralogy department at the University of Oslo by H. Hougen, E. Klüver, and O.A. Lökke (1925). Of the analyses published in that work, 78 are related to the clay of southern Norway. These 78 analyses were used for an average calculation. For this calculation, 68 analyses were of the primary zone of the relevant clay and deposits, and 10 analyses were of the secondary zone, in order to determine the extent to which weather-related phenomena appear after the deposits of the sediments. One observes such weather-related phenomena here, since the amount of the ferric (trivalent) iron in the secondary zone increased through oxidation, while the amount of calcium had been lessened through the leaching of carbonic lime. The entire picture shows great similarity with the figures from Clarke and Washington, an indication that the figures produced by these authors actually corresponds closely to the average of the magmatic silicate rock, despite the initial, justifiable objections against their calculation methods.

Recently, H. L. Vogt (1931) has issued critique of the figures from Clarke and Washington, on the basis of the argument that in their calculations too much consideration is given to rare types of rock. Vogt calculates the average compositions on the basis of Daly’s averages (1910) for the individual types of volcanic rocks, to which he gives a weight proportional to the mass of the relevant rocks. He gives two alternatives as a frequency ratio of the plutonic rock, assigning the granite either 50 or 60% of the total weight.

Proportion of Volcanic Rock
	I	II
Granite	50	60
Quartz monzonite and granodiorite	10	9
Quartz diorite and diorite	8	6
Gabbro	18	15
Anorthosite	4	3
Pyroxenite and peridotite	½	¼
Nordmarkite and pulaskite	1	1
Alkali-Lime-Syenite	3	2
Monzonite	4	3
Nepheline syenite	1	½
Essexite etc.	½	¼
	100	100

     

Average Composition of Volcanic Rocks and Norwegian Clay
	Clarke-Washington, 1924	J.H.L. Vogt, 1931		Norwegian Clay
		I	II	Primary zone	Secondary zone	Middle
SiO2	59.12	64.03	65.73	58.94	60.82	59.19
TiO2	1.05	0.600	0.546	0.79	0.78	0.79
Al2O3	15.34	15.71	15.41	15.87	15.48	15.82
Fe2O3	3.08	2.20	2.10	3.28	4.35	3.41
FeO	3.80	2.66	2.30	3.69	2.82	3.58
MnO	0.12	--	--	0.10	0.12	0.11
MgO	3.49	2.67	2.23	3.33	3.02	3.30
CaO	5.08	4.62	4.01	3.19	2.25	3.07
Na2O	3.84	3.51	3.43	2.05	2.00	2.05
K2O	3.13	3.52	3.79	3.95	3.83	3.93
H2O	1.15	--	--	3.01	3.10	3.02
P2O5	0.30	0.18	0.17	0.21	0.23	0.22
SO3	--	--	--	0.09	0.05	0.08
S	0.05	--	--	0.08	0.03	0.07
CO2	0.10	--	--	0.60	0.18	0.54

In the comparison of the Norwegian clay with the calculated average compositions of the volcanic rock, it becomes apparent that only the contents of calcium and sodium in the clays have experienced a considerable decrease as a result of the events of weathering and leaching. One notices further, that the leaching in the weathered secondary zone of the clay deposits is already somewhat progressed. The leaching of calcium and sodium by the weathering events and the forming of clayey and sandy sediments is one of the most important geochemical phenomena of the external cycle of these elements. In the Norwegian glacial and post-glacial clays this phenomenon appears in any case much weaker than in normal clayey sediments, because erosion and re-sedimentation happen so quickly and at such low temperatures that the chemical weathering could only occur limitedly. For this reason this clay would be suited for a calculation of the average composition of the lithosphere.

One plutonic rock, whose chemical composition corresponds closely to the average of volcanic rock, is the opdalite from the 150 km2 mass of plutonic rock of Opdal-Indset at approximately 62° 40’ N in southern Norway (Goldschmidt, 1916). Opdalite* is found primarily near the edges of the masses of plutonic rock, whose composition changes from quartz-biotite-norite to acidic, quartz-rich trondhjemite. The following table shows the chemical content of opdalite:

	I	II
SiO2	62.25	61.64
TiO2	0.94	0.97
Al2O3	15.15	15.44
Fe2O3	0.96	0.92
FeO	4.49	4.64
MnO	0.07	not determined
MgO	3.92	4.28
CaO	4.47	4.85
Na2O	0.06	not determined
K2O	3.30	3.55
H2O	3.50	3.24
P2O5	0.16	0.15
CO2	0.06	0.12
S	0.04	n. best.
H2O - 105°	0.05	n. best.
H2O 5°	0.57	0.43
	99.99	100.23
Density 20°/4°	2.777	2.790

*- Opdalite is an outdated term to describe a local variety of quartz diorite with biotite, amphibole and two pyroxene as mafic minerals. 

The following table gives two calculations of the mineral content according to the observed minerals, which are compared in column III with the “norm” of the average composition of the volcanic rock from Washington (1910, no reference?).

	I	II	III
Quartz	16.0	14.0	8.0
Potash feldspar	15.0	13.0	19.8
Albite	28.0	30.1	34.5
Anorthite	15.0	16.0	15.2
Diopside augite	4.7	5.2	7.7
Hypersthene	8.8	9.3	11.1
Magnetite	0.5	0.5	2.7
Ilmenite	1.2	1.2	0.9
Biotite	10.0	10.5	--
Apatite	0.4	0.4	--
Pyrrhotite	0.1	--	--
Calcite	0.1	0.3	--
	99.8	100.5	99.9

REFERENCES

1. Sederholm, J.J. (1925). The average composition of the Earth’s crust in Finland. Fennia 45, 18.
2. Hougen, H., Klüver, E. and Lökke O.A. (1925). Ündersokelser over norske lerer V. Statens Raastoffkom. Publ., 22, 1.
3. Vogt, J.H.L. (1931). On the average composition of the Earth’s crust, with particular reference to the contents of phosphoric and titanic acid. Skrifter utgitt av det Norske Vidensk. Akademi I Oslo, Mat.-naturv, 7.
4. Daly, R.A. (1910). Average chemical compositions of igneous-rock types. Proc. Am. Academy of Arts and Sciences, 45, 211.
5. Goldschmidt, V.M. (1916). Geologisch-petrographische Studien im Hochgebirge des südlichen Norwegens. IV. Übersicht der Eruptivgesteine im kaledoniscen Gebirge zwischen Stavanger und Trondhjem. Vid. Selsk. Skr. I, Mat.-naturv. 2.

REFERENCES FOR THE INTRODUCTION

1. Clarke, F.W. and Washington, H.S. (1924). The composition of the Earth's crust. United States Geological Survey, Professional Paper 127, 117.
2. Condie, K.,C. (1993). Chemical composition and evolution of the upper continental crust: constraining result from surface samples and shales. Chemical Geology,104, 1-37.
3. McDonough, W.F. and Sun, S.-s. (1995). The composition of the Earth. Chemical Geology, 120, 223-253.
4. Ronov, A.B. and Yaroshevsky, A.A., (1969). Chemical composition of the earth’s crust. In: The Earth’s Crust and Upper Mantle, AGU, Washingon D.C. 13, 2-7.
5. Rudnick, R.L. and Fountain, D.M. (1995). Nature and composition of the continental crust: a lower crustal perspective. Reviews in Geophysics, 33, 267-309.
6. Taylor, S.R. and McLennan, S.M. (1985). The continental crust: its composition and evolution. Blackwell Scientific Pub., Palo Alto, California.
7. Turekian, K.K. and Wedepohl, K.H. (1961). Distribution of the Elements in some major units of the Earth's crust. Geological Society of America, Bulletin, 72, 175-192.
8. Vinogradov, A.P. (1962). Average contents of chemical elements in the principal types of igneous rocks of the Earth's crust. Geochemistry Int., 7, 641-664.