The constituents of kimberlite are commonly divided into two genetic groups of proper kimberlite minerals and those produced by disintegration of crustal and/or mantle substrates. However, surprisingly, no discrimination criteria have been so far developed. The intratelluric (or xenogenic) origin of a mineral can be either inferred implicitly or presumed to come from the mantle as “everything in kimberlite” does. Yet, statistical composition relationships of a mineral with its host can provide unambiguous evidence for its origin (Minin et al., 2010).

We in this study correlate the compositions of pyrope and the host kimberlite in samples from the Yakutian diamond province.

Kimberlite differs from other igneous rocks as it encloses abundant xenolithic material its parent melt entraps on the way to the upper lithosphere and at rooting sites of volcanic pipes (diatremes). Describing kimberlite in terms of igneous geology and petrological reconstructions requires looking into its groundmass freed from xenoliths. Technologically, the problem of separating the impurity components was solved in the following way. A piece of core or a rock sample were cut into several plates; each xenolith and autolith in a plate were contoured (or filled if necessary) with a felt marker; then the marked xenoliths and autoliths were hand-picked under a microscope after the plate was powdered (with a mortar or an anvil), and were removed; finally, the groundmass was re-examined and passed further to chemical or mineralogical analyses.

The horizontal and vertical distribution of major oxides in kimberlite pipes showed highly variable patterns: they are 20 to 50% higher or lower than the average within distances of one or two meters. Similar compositional variations appear also in pyropes from the same kimberlites. Given this variability, the only way to recognize the trends of composition change in kimberlite and pyrope is obviously to group the data, specifically, to apply averaging over a pipe. This approach is justified by the law of large numbers implying that the arithmetic sample mean of quite many random numbers differs infinitely little from its mathematic expectation to a probability arbitrarily close to the unity.

The compositions of Yakutian kimberlites used for correlation were from published evidence (Vasilenko et al. (2006). The pipe-average compositions of kimberlite-hosted pyropes were calculated from our own data, from unpublished reports of the Botuobia Geological Survey, and from data collected by I. Ashchepkov. Altogether we processed 6749 kimberlite compositions and 3943 pyrope compositions from 33 pipes in the Upper Muna, Daldyn, Alakit-Markha, Nakyn, and Mirnyi fields. The compositions of pyropes and their hosts show quite a good correlation (Table 1). The contents of Ti and Ca oxides in kimberlite increase proportionally to those in pyrope. The contents of Cr₂O₃ and MnO in pyropes are inversely proportional to TiO₂ in kimberlite. A similar relationship appears between CaO in pyrope and MgO in kimberlite, while MgO in pyrope is directly proportional to P₂O₅ in kimberlite.

Table 1. Pyrope-kimberlite correlation coefficients (n=33, r_0.01=0.43) for major oxides (pipe bulk average), Yakutian diamond province

Major oxides in kimberlite	Major oxides in pyrope
Major oxides in kimberlite	TiO₂	Cr₂O₃	MnO	MgO	CaO
TiO₂	0.43	-0.48	-0.49
MgO					-0.48
CaO					0.47
P₂O₅				0.44

The pyrope-kimberlite compositional correlation is observed for major oxides present in both the rock and the enclosed mineral, as well as for those absent from the latter (e.g., P₂O₅). This may happen only if the mineral crystallizes from magma which is parent for the rock as a whole. More evidence for magmatic crystallization of pyrope comes from the straight form and uniform filling of regression lines in the variation diagrams (Fig. 1).

It is known from experimental petrology (Doroshev et al., 1997) that Cr contents in pyrope are a function of pressure: higher-Cr pyropes crystallize at high pressures. The negative correlation between Cr₂O₃in pyrope and TiO₂ in kimberlite (Table 1 and Fig. 1) provides further justification for the use of the Cr oxide as reference in classifying kimberlites according to origin depths and, hence, according to diamond productivity (Vasilenko et al.. 1997).

Doroshev A.M., Brai G.L., Girnis A.V., Turkin A.I., Kogarko L.N., 1997. Pyrope-knorringite pyropes in the Earth’s mantle: an experimental study in the system MgO-Al2O3-SiO2-Cr2O3. Geologiya i Geofizika (Russian Geology and Geophysics), 38 (2), 523-545.

Minin V.A., Prugov V.P., Podgornykh N.M., Kovyazin S.V., Holodova L.D, 2010. Compositional features of kimberlite-hosted chromites from the Botuobia pipe (Yakutia). Reports, Rus. Mineral. Soc., 139 (3), 54-72.

Vasilenko V.B.; Zinchuk N.N.; Kuznetsova L.G., 1997. Major-Element Composition Models of Diamond Deposits in Yakutia [in Russian]. Nauka, Novosibirsk, 557 pp.

Vasilenko V.B., Zinchuk N.N., Kuznetsova L.G., Minin V.A., Kholodova L.D., 2006. Average compositions of kimberlite pipes in the Vlyui subprovince of Yakutia. Vestn. Voronezh University, Ser. Geol., No. 2, 126-140