Abstract– Xenoliths are inclusions of a given meteorite group embedded in host meteorites of a different group. Xenoliths with dimensions between a few μm and about 1 mm (microxenoliths) are “meteorite-trapped” analogues of micrometeorites collected on the Earth. However, they have the unique features of sampling the zodiacal cloud (1) at more ancient times than those sampled by micrometeorites and (2) at larger distances from the Sun (corresponding to the asteroid Main Belt) than that sampled by micrometeorites (1 AU). Herein we describe a systematic search for new xenoliths and microxenoliths in H chondrites, aimed at determining their abundance in these ordinary chondrites, analyzing their mineralogy, and searching for possible correlations with host meteorite properties. Sixty-six sections from 40 meteorites have been analyzed. Twenty-four new xenoliths have been discovered. About 87% of them are microxenoliths (i.e., <1 mm), only three are >1 mm in their largest dimension. All the newly discovered xenoliths and microxenoliths are composed of carbonaceous chondritic material. Hence, the zodiacal cloud was dominated by carbonaceous material even in past epochs. All the new xenoliths and microxenoliths have been found in regolith breccias. Hydrous-phase-rich xenoliths and microxenoliths in H4 and H5 chondrites attest that their embedding happened after the end of the thermal metamorphism. All these data suggest that xenoliths and microxenoliths were embedded when their host meteorites were part of the parent body regolith. This, combined with the H chondrite impact age distribution, attests that the embedding may have happened as early as 3.5 Gyr ago.
Xenoliths have long been recognized as an important subject in meteoritics (see e.g., Wahl 1952). The word xenoliths refers to “inclusions in a rock to which they are not genetically related” (Keil 1982). As most meteorites are formed by agglomeration of components with different origins, the word xenolith more precisely refers to inclusions of a given group or class embedded in host meteorites of a different group or class. Therefore, xenoliths are distinct from cognate clasts (fragments related to the host rock) and, more generally, from meteorite constituents named clasts.
The study of xenoliths sheds light on different issues concerning meteorites and the solar system. For instance, xenoliths can help to understand the history of relationships between different meteorite parent bodies (see e.g., Wilkening 1977); or they could reveal new classes of parent bodies, namely they could be samples of distant solar system bodies such as comets and Kuiper Belt objects, for which at present we do not have certain meteoritic samples (Campins and Swindle 1998; Gounelle et al. 2006, 2008; Zolensky et al. 2009; Gounelle 2011).
Xenoliths that have sizes between a few μm and 1 mm suggest relations with micrometeoroids in the zodiacal cloud (see e.g., Grün et al. 2004) and with micrometeorites recovered on Earth (see e.g., Genge et al. 2008). Micrometeoroids are, numerically, the most abundant objects in the inner solar system and their size distribution is dominated by 200 μm-size micrometeoroids (Grün et al. 1985; Nesvorný et al. 2010). Micrometeorites are usually referred to as extraterrestrial particles with sizes between a few μm and about 1 mm (Flynn et al. 1993; Genge et al. 2008). Herein, by analogy with the cases of micrometeoroids and micrometeorites, we use the word microxenoliths to indicate meteorite xenoliths with maximum dimension <1 mm.
Gounelle et al. (2003) pointed out that microxenoliths are not fragments of larger objects, but real micrometeoroids (i.e., zodiacal cloud dust) embedded in meteorite parent bodies. They observed that in howardites the population of microxenoliths shows different relative abundances of CM2 and CR2 material with respect to the population of larger xenoliths observed in the same meteorites by Zolensky et al. (1996). Also, Gounelle et al. (2003) pointed out that in the inner Solar System the probability of impact on a meteorite parent body is a factor of approximately 10 larger for micrometeoroid-sized objects than for meteoroid-sized objects (Grün et al. 1985; Halliday et al. 1989; Love and Brownlee 1993; Bland et al. 1996). A critical condition for carbonaceous micrometeoroids to become hydrated, unshocked microxenoliths is that their impact with the host meteorite parent body must happen at very low velocity, of the order of a few hundred m s−1 (Rubin and Bottke 2009). Numerical simulations by Gounelle et al. (2003) showed that about 10% of asteroidal micrometeoroids impact the asteroid Vesta with a velocity <1 km s−1. Also, numerical simulations by Briani et al. (2011) showed that, considering only impact velocities <1.5 km s−1, 200 μm-diameter micrometeoroids of both asteroidal and cometary origin collide with possible H chondrite parent bodies with mean impact velocity of approximately 350 m s−1.
Microxenoliths are the most similar objects to micrometeorites (Gounelle et al. 2003), i.e., the most abundant material collected by Earth from space (Love and Brownlee 1993). However, since micrometeorites are destroyed by the Earth’s geological activity, microxenoliths sample more ancient epochs of the zodiacal cloud than micrometeorites. At the same time, having been embedded in meteorite parent bodies, microxenoliths represent the zodiacal cloud dust that resides in the region of the asteroid Main Belt, i.e., between 2 and 4 AU, or even farther, while micrometeorites are samples of the zodiacal cloud at 1 AU.
In search of new xenoliths and microxenoliths, we chose to concentrate our efforts on H chondrites because, on the basis of previous works, xenoliths and microxenoliths appear to be more abundant in H chondrites than in other ordinary chondrites and achondrites (with the exception of howardites, see Table 1). However, a systematic search for and characterization of xenoliths and microxenoliths has been previously performed only for howardites (Zolensky et al. 1996; Gounelle et al. 2003, 2005). Hydrated and unshocked xenoliths have already been found in H chondrites (see e.g., Rubin and Bottke 2009). These are the xenoliths and microxenoliths that least suffered from shocks during the embedding process in the host meteorite, and therefore they are the most representative of their parent bodies.
Several meteorites present in this table also contain noncarbonaceous chondrite xenoliths, not reported here (see e.g., Bischoff et al. 2006).
Meteorite name abbreviations: ALH = Allan Hills (Antarctica); DaG = Dar al Gani (Libya); EET = Elephant Moraine (Antarctica); HaH = Hammadah al Hamra (Libya); LEW = Lewis Cliff (Antarctica); NWA = Northwest Africa; PAT = Patuxent Range (Antarctica); QUE = Queen Alexandra Range (Antarctica).
Our goals are to investigate the abundance and size distribution of xenoliths and microxenoliths in H chondrites, to verify if carbonaceous chondritic xenoliths and microxenoliths are really common in H chondrites, as it appears from previous works (see Table 1 for references), to search for xenoliths and microxenoliths related to meteorites other than carbonaceous chondrites and for possible “exotic” xenoliths and microxenoliths, clearly different from known meteorite classes. Also, we investigate the relations between xenoliths and their host meteorites, to understand when xenoliths and microxenoliths were embedded in the H chondrite parent body.
The question about the origin of xenoliths and microxenoliths in H chondrites is investigated in more detail in a companion article (Briani et al. 2011), where we describe dynamical simulations aimed at estimating the contribution of primordial asteroids and Jupiter family comets as possible parent bodies of microxenoliths.
Table 1 shows that the most carbonaceous chondrite xenolith-rich classes among meteorites are ordinary chondrites and the HED group. Among ordinary chondrites, H chondrites are the group where the most carbonaceous chondrite xenoliths have been found. However, descriptions of more than two or three xenoliths in the same meteorite are rare. In contrast, howardites are very rich in carbonaceous chondrite xenoliths: considering seven meteorites (Kapoeta, Bholgati, Jodzie, Mundrabilla 020, LEW 87015, LEW 85441, and Yamato 793497), Zolensky et al. (1996) reported 72 xenoliths. Most important, howardites are also rich in xenoliths of size <1 mm, i.e., in microxenoliths. Gounelle et al. (2003) described 61 microxenoliths found in Kapoeta, Jodzie, and Yamato 793497. For howardites, millimeter-sized carbonaceous chondrite xenoliths appear to be different from the smaller microxenoliths found in the same meteorites. Among carbonaceous chondrite xenoliths with size >1 mm, most are composed of CM2 material, with a CR2 to CM2 xenolith ratio of 0.18 (Zolensky et al. 1996). In contrast CM2 and CR2 microxenoliths are observed in almost the same abundances, with CR2/CM2 = 1.1 (Gounelle et al. 2003). This suggests that carbonaceous chondrite microxenoliths are not fragments of larger objects, as xenoliths larger than 1 mm are, but real micrometeoroids embedded in the host meteorites.
Materials and Methods
H Chondrite Samples
The collection of H chondrite sections prepared in the Laboratoire de Minéralogie et Cosmochimie du Muséum at the Muséum National d’Histoire Naturelle in Paris consists of more than 130 samples that are polished blocks or thin sections. Samples include petrologic types from 3 to 6, as well as samples of different shock degree, different cosmic-ray exposure age, samples of brecciated meteorites, and of solar gas-rich meteorites.
We prepared 15 new sections of H chondrite regolith breccias from whole rock samples of Leighton, Plainview 1917, Tysnes Island, Bremervörde, Weston, and Dimmitt. We cut each sample with a circular diamond saw, then we embedded the piece in KörapoxTM epoxy (particularly suited for high vacuum analyses) and we polished it using aerosols with diamonds of different sizes, from 3 μm down to a 0.25 μm finish.
In addition, we analyzed a few polished thin sections from the collection of the Astromaterial Research and Exploration Science laboratory of the NASA Johnson Space Center (Houston, Texas). We selected a few sections in which xenoliths have previously been identified (Zolensky et al. 2009). These sections include samples of Willard (b) and Abbott. Also a section of Sharps from the Smithsonian Institution (USNM 640) was studied.
A list of the samples we analyzed is reported in Table 2. Most of the analyzed meteorites are falls, and the terrestrial weathering grade for terrestrial finds, when known, is quite low.
Table 2. List of H chondrite sections analyzed in this work.
Find or fall
Analyzed surface (mm2)
aTerrestrial weathering data are from MetBase v7.2. Weathering categories A, B, and C are assigned to hand specimens by the Meteorite Working Group at the NASA Johnson Space Center (A: minor rustiness; B: moderate rustiness; C: severe rustiness). Weathering categories W0 to W6 are defined in Wlotzka (1993).
MNHN-1, MNHN-2009_06, _07
Grady 1937 (H3.7)
Praire Dog Creek (H3.7)
MNHN-1, MNHN-2, USNM640
Willard (b) (H3.6)
Total surface analyzed for H3 chondrites =
MNHN-1, MNHN-2, MNHN-3
MNHN-2009_01, _02, _03, _04, _05
MNHN-2007_24, _25, _26
Total surface analyzed for H4 chondrites =
Total surface analyzed for H5 chondrites =
Total surface analyzed for H6 chondrites =
Mixed H chondrites
Sahara 00181 (H4–6)
Sahara 99070 (H3.8-6)
MNHN-1, MNHN-3867-1, MNHN-3867-2
Sahara 97179 (H5–6)
Tulia (a) (H3–4)
MNHN-1, MNHN-2, MNHN-3
Total surface analyzed for mixed H chondrites =
Searching for Xenoliths and Microxenoliths: Instruments and Techniques
Our systematic search for new xenoliths and microxenoliths started with an examination of samples by optical microscopy. Meteorite polished blocks and thin sections were scanned in reflected light. We limited our search to inclusions larger than 50 μm in their maximum dimension. The most visible characteristic indicating that an inclusion is a potential xenolith is the presence of fine-grained matrix. A possible limit of this approach is that identification based on the presence of fine-grained matrix makes it easier to detect possible carbonaceous chondrite xenoliths and microxenoliths than possible ordinary chondritic xenoliths and microxenoliths, because carbonaceous chondritic material generally has higher matrix abundances than ordinary chondritic material. For type 4 to 6 H chondrites, potential xenoliths and microxenoliths can be easily distinguished from the meteorite matrix, because in these meteorites the matrix is generally composed of recrystallized grains (Huss et al. 1981; Weisberg et al. 2006). For less metamorphosed, type 3 H chondrites, it is often more difficult to distinguish xenoliths and microxenoliths from the host meteorite matrix. However, matrix is usually dispersed among the meteorite components or forms particular structures (e.g., chondrule rims), while xenoliths and microxenoliths often have sharp edges. We found only one xenolith in an H3 chondrite (Willard b): in this case the xenolith size and its petrographic properties clearly show that it is a CM-like inclusion.
To ascertain whether selected inclusions are xenoliths or not, scanning electron microscopy (SEM) images were acquired. Energy dispersive X-ray (EDX) spectroscopy was performed to identify the different phases present. This technique is also suitable to identify possible ordinary chondritic xenoliths and microxenoliths in H chondrites, by searching for different chemical compositions (e.g., olivine and pyroxene grains with elevated Fa and Fs values, respectively, could indicate the presence of L- or LL-type xenoliths). SEM and EDX spectral analyses have been performed at the Laboratoire de Minéralogie et Cosmochimie du Muséum of the Muséum National d’Histoire Naturelle (Paris, France) by means of a JEOL JSM 840-A SEM equipped with an EDAX Genesis EDX detector. EDX spectral maps were obtained for some xenoliths.
In a few cases, high resolution SEM images were acquired at the University of Paris VI using a Zeiss Supra-55 VP field emission gun scanning electron microscope. With this instrument we were able to perform EDX spectral analyses of submicrometer-sized grains in the matrix of a few xenoliths.
Quantitative analyses for the mineralogical composition of xenolith matrix and components were obtained with a CAMECA SX-50 electron microprobe (EMP) at the University of Paris VI. A 10 nA focused beam, at a 15 kV acceleration voltage, was used for point analyses of matrix as well as for larger silicates, oxides, sulfides, and metals. The beam diameter was about 1 μm. The mean composition of the fine-grained matrix was obtained by averaging several (≥7) point analyses for each xenolith and microxenolith. For matrix analyses, data with values of S ≥ 5 wt%, which cannot be interpreted as due to tochilinite, were considered as due to contamination by a nearby sulfide (typically larger than matrix grains) and hence corrected assuming the sulfide is troilite (i.e., all the S is attributed to the troilite and the appropriate amount of Fe is subtracted from the data). For analyses of carbonates we used a 4 nA defocused beam at a 15 kV acceleration voltage. In each case natural and synthetic standards were used for calibration. Detection limits were always lower than 2000 ppm.
Classification of Xenoliths and Microxenoliths
Xenoliths that have been described previously have been mainly classified on the basis of their mineralogical, structural, and compositional similarities with known groups of meteorites. Therefore, carbonaceous chondrite xenoliths are generally referred to as CI-like, CR-like, or CM-like. However, classifying xenoliths and microxenoliths on the basis of their similarities with known groups of meteorites presents some ambiguities, because some xenoliths are different from known carbonaceous chondrite material. Indeed differences remain between xenoliths and carbonaceous chondrites. For xenoliths in CB and CH chondrites described by Greshake et al. (2002), no unambiguous assignment to any known group of carbonaceous chondrites has been established, even if these xenoliths clearly have close relationships with C1 and C2 material (such as matrix mineralogy, magnetite morphology, and chemical composition of major phases). Gounelle et al. (2003) reported differences between CR2-like microxenoliths in howardites and CR2 chondrites. Microxenoliths in howardites have abundant matrix, highly fayalitic olivine grains, and Fe-rich saponite, while calcium-aluminum-rich inclusions (CAIs) and chondrules are absent and metal is rare. Minor differences are also reported for CM2-like microxenoliths in howardites with respect to CM2 chondrites: in microxenoliths the pyroxene to olivine ratio is higher, saponite is more abundant, while chondrules and Mg-rich olivine are less abundant. Rubin and Bottke (2009) described mineralogical and compositional differences between CM-like xenoliths in Abbott and CM chondrites, such as differences in the sulfide modal abundance and in the S/SiO2 ratio for tochilinite–cronstedtite.
To solve this problem, alternative classification schemes have been proposed, directly based on observed properties of xenoliths. Gounelle et al. (2005) proposed three groups for microxenoliths in howardites, defined by the presence or absence of tochilinite, magnetite, and olivine. For comparison with the previous classification scheme, tochilinite-rich, magnetite-poor microxenoliths are those most similar to CM chondrites. Tochilinite-poor, magnetite-rich microxenoliths are subdivided into two more groups, based on the abundance of olivine. Olivine-rich microxenoliths are those most similar to CR chondrites, while olivine-poor microxenoliths are CI-like.
Similarly, considering the presence of magnetite and anhydrous silicates as key reference properties, we divided the newly discovered xenoliths and microxenoliths into three groups: “magnetite-rich, hydrous,”“magnetite-poor, semi-hydrous,” and “magnetite-rich, semi-hydrous.” We do not make associations between the newly observed xenoliths and microxenoliths and carbonaceous chondrite groups. Rather, we compare our observation to data that have been previously reported for xenoliths and microxenoliths in other meteorites.
Results: Mineralogy and Petrography
We analyzed 66 sections from 40 different meteorites (Table 2). The total surface analyzed is 95.22 cm2 (in Table 2 the surface studied for each section and each H chondrite petrologic type is reported). We discovered 24 new xenoliths. Among these, only three have a maximum dimension larger than 1 mm, i.e., about 87% are microxenoliths. We report our results starting with microxenoliths, classified into three groups on the basis of their magnetite and anhydrous silicate content, plus one peculiar microxenolith found in Leighton (H5). We then describe three new greater mm-sized xenoliths. The main properties of all the newly discovered xenoliths and microxenoliths are reported in Table 3. EMP data for point analyses of xenoliths and microxenoliths matrix and anhydrous silicates are reported in Table 4.
Table 3. Properties of xenoliths and microxenoliths identified in this studya.
As microxenoliths have irregular shapes, we characterize them by measuring (1) their maximum dimension and (2) the maximum dimension in a direction perpendicular to the first measurement. This leads to the size distribution shown in Fig. 1: more than half of the microxenoliths have maximum dimensions ≤200 × 200 μm. The mean microxenolith maximum dimension is 209 μm.
Magnetite-Rich, Hydrous Microxenoliths
Ten microxenoliths contain magnetite and no anhydrous silicates (Figs. 2A and 2B). They have a maximum dimension ranging between 60 and 310 μm. All these microxenoliths are dominated by a fine-grained matrix. Low analytical totals (about 83%) obtained in EMP analyses suggest that the matrix is composed of phyllosilicates. Also matrix porosity certainly contributes to low analytical totals. However, our results are similar to those obtained for phyllosilicates in carbonaceous chondrite xenoliths and microxenoliths of howardites (Zolensky et al. 1996; Gounelle et al. 2003). Most matrix point analyses, plotted in an atomic Mg-Fe-(Si + Al) ternary diagram (Fig. 3A), lie between the saponite and serpentine solid solution lines, some points are below the serpentine solid solution line (within the CM2 serpentine field) with a few Fe-rich (Fe > 60 atom%) points. Magnetite is present as spheroidal and hexagonal micrometer-sized grains and as smaller, submicrometer-sized framboidal grains. No anhydrous silicates or metal grains were observed. Sulfides grains are mainly pyrrhotite, with Ni contents around 2 wt%; rare pentlandite grains were observed. Carbonates are generally absent, but a few dolomite and breunnerite grains were observed in four microxenoliths (see e.g., Fig. 2B). The main properties of these microxenoliths are similar to those of xenoliths and microxenoliths that were described as CI-like in previous studies (Zolensky et al. 1989; Brearley and Prinz 1992; Buchanan et al. 1993; Gounelle et al. 2003; Nakashima et al. 2003).
Magnetite-Poor, Semi-Hydrous Microxenoliths
We identified three microxenoliths that contain anhydrous silicates and little or no magnetite (Figs. 2C and 2D). They have maximum dimensions ranging between 100 and 270 μm. They are dominated by a fine-grained matrix, for which EMP point analyses show low analytical totals (approximately 83%), suggesting the presence of phyllosilicates and a porous structure. In an atomic Mg-Fe-(Si + Al) ternary diagram (Fig. 3B) EMP data plot mostly below or close to the serpentine solid solution line (in the CM2 serpentine field), with some above the serpentine line (in the CR2 phyllosilicates field). Matrix analyses are rich in S (up to 4 wt%), probably due to the presence of submicrometer-sized sulfides, such as those observed in the matrix of CR3 chondrites (Abreu and Brearley 2010). Olivine and pyroxene grains are common, their sizes ranging from a few μm up to 50 μm. They are mainly Mg-rich, with olivine compositions ranging between Fo74 and Fo82 (mean = Fo80), and low-Ca pyroxene composition ranging between En60Wo4, En93Wo1, and En68Wo6. Two Ca-rich pyroxene grains, with compositions En69Wo21 and En67Wo23, respectively, were observed in microxenolith #4 in Leighton 2007_22. In the same microxenolith two low-K plagioclase (Ab92) grains are present. The sulfides troilite and pyrrhotite are present. Fe-Ni metal grains with sizes of a few μm are present. Most of them are kamacite, but some are very rich in Ni (up to 50 wt%, probably tetrataenite). Carbonates are not abundant, but we observed a couple of grains per microxenolith, with sizes up to 20 μm. They are Ca-carbonates and breunnerite. For howardite microxenoliths rich in anhydrous silicates and poor in magnetite, Gounelle et al. (2003) proposed an association with CM2 chondrites. However, the H chondrite magnetite-poor, semi-hydrous microxenoliths described herein differ in one important respect: EMP analyses did not reveal the presence of tochilinite in the fine-grained matrix. Tochilinite is usually considered as a signature of CM-like material. In addition Fe-Ni metal grains are fairly common in these microxenoliths. This suggests affinities with CR-like material. However, these microxenoliths lack magnetite, which is generally common in CR chondrites (with the exception of unaltered CR3 chondrites, Abreu and Brearley 2010).
Magnetite-Rich, Semi-Hydrous Microxenoliths
Seven microxenoliths are rich in both magnetite and anhydrous silicates (Figs. 2E and 2F). Their sizes range between 36 and 575 μm and they are dominated by fine-grained matrix. In an atomic Mg-Fe-(Si + Al) ternary diagram (Fig. 3C) EMP analyses of matrix mostly lie above or close to the serpentine solid solution line (in the CR2 phyllosilicate field), with some points lying below the serpentine line (in the CM2 serpentine field). Matrix point analyses have low analytical totals (about 83%), and show high contents of S (up to 5 wt%), probably due to submicrometer-sized grains of sulfides (Abreu and Brearley 2010). Olivine and pyroxene grains are somewhat less abundant than in magnetite-poor, semi-hydrous microxenoliths. Olivine compositions range between Fo68 and Fo99 (the mean is Fo91). Pyroxene grains are also Mg-rich, their compositions ranging between En87Wo4 and En96Wo2. One ilmenite grain was observed in microxenolith #11 in Leighton 2007_21. Sulfides are common, mainly pyrrhotite grains up to a few μm in size, and less frequently troilite grains. We did not observe Fe-Ni metal grains. Calcium-carbonates and dolomite grains are more abundant than in magnetite-poor, semi-hydrous microxenoliths, although in several microxenoliths no carbonates were observed. Microxenolith #2 in Leighton 2007_22 contains several calcium-carbonate grains and abundant magnetite in the form of both aggregates of submicrometer-sized framboidal grains and larger spheroidal grains (Fig. 2F). Microxenolith 3a in the same section contains a dolomite grain 30 μm across. Previous studies associated xenoliths (Zolensky et al. 1996) and microxenoliths (Gounelle et al. 2003) that contain both anhydrous silicates and magnetite (as this group of H chondrite microxenoliths) with CR chondrites.
Microxenolith #1 in section 2007_22 of Leighton (Fig. 4A) has unique properties. The absence of anhydrous silicates and the presence of pyrrhotite grains and framboidal magnetite aggregates would suggest that this microxenolith is similar to CI-like microxenoliths (Nakashima et al. 2003). However, it contains a high abundance of carbonate grains, rarely observed in CI-like microxenoliths. These carbonates are dolomite grains with dimensions ranging from approximately 1 μm to greater than 10 μm (Fig. 4B). Their abundance resembles that observed in the carbonate-rich lithology of Tagish Lake (Zolensky et al. 2002), but carbonates in the Leighton microxenolith do not show the complex structure (Ca core + Mn-Fe-Mg rims) of the Tagish Lake carbonates. In addition this microxenolith does not have the same characteristics of the carbonate-rich lithology of Tagish Lake: the matrix composition is that of CM2 serpentine, while the carbonate-rich lithology of Tagish Lake contains mostly saponite. Magnetite is rare in the carbonate-rich lithology of Tagish Lake, but is abundant in the Leighton microxenolith. Dolomites of xenolith #1 in Leighton 2007_22 contain 24.4–27.6 wt% CaO, 16.4–19.9 wt% MgO, 2.5–4.2 wt% FeO, and 2.8–6.1 wt% MnO. This composition is similar to that of CI and CM dolomites (Fig. 5) and to that of dolomites found in Kaidun (Weisberg et al. 1994).
In the same polished section of Leighton there is the magnetite-rich, semi-hydrous microxenolith #2 that has similar size (575 × 300 μm), but contains olivine and pyroxene grains, abundant framboidal magnetite aggregates, and carbonate grains (dolomite and Ca-carbonates). This suggests also that dolomite grains of xenolith #1 in Leighton 2007_22 are not due to terrestrial aqueous alteration, otherwise they should be abundant also in microxenolith #2 and in the other microxenoliths found in the same section (microxenoliths 3a, 3b, and #4 have very few carbonates, some of which are breunnerite).
Xenoliths (Size >1 mm)
Two xenoliths with maximum size >1 mm were identified in Abbott and one in Willard (b). These three xenoliths have mineralogical and petrologic similarities to CM-like carbonaceous chondrite xenoliths described in previous works (Zolensky et al. 1996; Rubin and Bottke 2009).
The two xenoliths discovered in Abbott (H3-6) are in the same polished section, which has a typical H6 lithology. Their dimensions are 5.4 × 4.6 mm and 2 × 1.5 mm.
The smaller (2 × 1.5 mm) xenolith in Abbott (Fig. 6A) is dominated by a fine-grained matrix, composed of serpentine and tochilinite–cronstedtite intergrowths (Table 5). Figure 7A shows the matrix composition in an atomic Mg-Fe-(Si + Al) ternary diagram: all points fall in the CM2 serpentine field. Sulfide grains are present within the matrix with compositions ranging from troilite to pyrrhotite and compositions intermediate between pyrrhotite and pentlandite (Ni content is up to 20 atom%), similar to compositions reported for many CM chondrites (Zolensky et al. 1993). No metal grains, magnetite, or carbonates were observed. The xenolith contains numerous rounded, Fe-poor areas composed mainly of fine-grained serpentine. We interpret these areas as relict chondrules, i.e., their phyllosilicates are the result of aqueous alteration of chondrule silicates. Embedded in the fine-grained matrix (Fig. 6B) are micrometer-sized sulfides and iron oxide grains. Such grains, as well as similar, diffuse veins with the same composition and the same low totals in EMP analyses, are probably terrestrial weathering products. Only three grains of olivine were found with grain sizes between 30 and 60 μm, their composition being very Mg-rich (from Fo96 to Fo99). Pyroxene grains have not been observed in this xenolith.
Table 5. Mean compositions of matrix and anhydrous silicates in CM-like xenoliths.
Abbott 7F larger xen
Abbott 7F smaller xen
Willard (b) 5 xen
Abbott 7F larger xen
Abbott 7F smaller xen
Willard (b) 5 xen
Abbott 7F larger xen
Abbott 7F smaller xen
Willard (b) 5 xen
Willard (b) 5 xen
Data in wt%. b.d. = below detection limit.
xen = xenolith.
No. of analyses
The larger (5.4 × 4.6 mm) xenolith in Abbott (Fig. 6C) is similar to the smaller one. Its matrix has the composition of CM2 serpentine (Table 5 and Fig. 7B). Intergrowths of tochilinite–cronstedtite are present (Table 5), but less abundant than in the smaller Abbott xenolith. Terrestrial weathering iron oxide grains and veins are abundant. Sulfide grains are common, with sizes up to 45 μm and compositions ranging from troilite to pyrrhotite as well as Ni-rich sulfides (Ni content is up to 28 atom%). Metal grains, magnetite, and carbonates have not been observed. However, this xenolith contains several chondrules, similar to porphyritic olivine chondrules, composed of olivine crystals, fine-grained serpentine and iron oxide alteration grains, always rimmed by fine-grained serpentine (Fig. 6D). Olivine grains are clearly more abundant than in the smaller xenolith. Including those in chondrules, their sizes range from a few μm up to 77 μm and their compositions are between Fo44 and Fo100, with a majority of grains being forsterite (22 of 36 grains have Fo>96). A large forsterite grain, 730 × 360 μm in size, is included in this xenolith (visible in the lower right part of the xenolith, Fig 6C). Once again pyroxene grains are absent.
The third xenolith (Fig. 6E) was found in the H3.6 chondrite Willard (b). This xenolith is 9 × 6 mm. Its fine-grained matrix is composed of serpentine and tochilinite–cronstedtite intergrowths (Table 5). All analyzed matrix points lie within or close to the CM2 serpentine range in an atomic Mg-Fe-(Si + Al) ternary diagram (Fig. 7C). Also present are Fe-poor areas composed of serpentine, probably produced by aqueous alteration of chondrule silicates, as in the Abbott xenoliths. These areas also form rims around relict chondrules, as in the larger Abbott xenolith. The Willard (b) xenolith (Fig. 6F) contains abundant olivine grains, with a wide size range (from a few μm to 200 μm). Olivine has a wide range of compositions, from Fo48 to Fo100, with the most common composition being Fo99 (30 of 41 grains have Fo98–100). Pyroxene grains are far less abundant (in a random survey, only three grains were identified). Pyroxene composition varies from En92Wo3 to En98Wo1. One spinel grain (MgAl2O4) is present in this xenolith, with an almost stoichiometric composition: 14.2 atom% Mg and 27.8 atom% Al. Abundant Ca-carbonate grains were observed in this xenolith. Sulfides are present as pyrrhotite and pentlandite, but no metal grains and no magnetite grains have been observed.
To classify the xenoliths and microxenoliths observed in H chondrites, we adopted a classification scheme based on the abundance of anhydrous silicates and magnetite. Among the 21 new microxenoliths, ten are magnetite-rich, hydrous; seven are magnetite-rich, semi-hydrous; three are magnetite-poor, semi-hydrous; and one is ungrouped. The three xenoliths (size >1 mm) are magnetite-poor, semi-hydrous, and similar to CM2 chondritic material. This classification highlights a fundamental result: carbonaceous chondritic xenoliths and microxenoliths are by far the most common in H chondrites. Our systematic search for new xenoliths and microxenoliths was not a priori biased toward this particular class of objects, but different xenoliths have not been found. Possibly ordinary chondritic xenoliths and microxenoliths (i.e., pieces of L or LL chondrites) are more difficult to identify than the carbonaceous ones, even if ordinary chondrite xenoliths in ordinary chondrites were previously reported (Rubin et al. 1983; Lipschutz et al. 1989; Bischoff et al. 2006). However, if ordinary chondritic xenoliths and microxenoliths were as abundant as the carbonaceous ones, then we should have found some of them.
A second important result is that the vast majority (about 87%) of the observed xenoliths and microxenoliths have maximum dimension <1 mm. As explained in the Introduction, we believe that microxenoliths are real zodiacal cloud micrometeoroids and not fragments of larger objects. Observations reported here are consistent with this assumption. First, because the observed microxenoliths have different compositions from the larger xenoliths. The latter are composed of CM-like material, with a matrix characterized by the presence of tochilinite, as also previously reported (Rubin and Bottke 2009). Instead, tochilinite was not observed in any of the microxenoliths. Second, because different microxenoliths were observed in single polished sections (as sections MNHN-2007_21 and MNHN-2007_22 of Leighton). Section MNHN-2007_21 of Leighton contains three magnetite-rich, semi-hydrous microxenoliths, two magnetite-poor, semi-hydrous microxenoliths, and one magnetite-rich, hydrous microxenolith. Such a concentration of different microxenoliths would be highly improbable if they were fragments of larger objects, given that for a meteorite parent body the probability of impact with submillimeter objects is a factor of about 10 larger than the probability of impact with a larger object (Grün et al. 1985; Halliday et al. 1989; Love and Brownlee 1993; Bland et al. 1996), and given that larger projectiles imply more violent impacts and therefore a lower probability of producing hydrated, unshocked fragments.
The prevalence of carbonaceous chondritic microxenoliths is consistent with the abundance of carbonaceous chondritic particles among micrometeorites collected on Earth. Micrometeorites represent the major part of the extraterrestrial matter flux that annually enters the Earth’s atmosphere (Love and Brownlee 1993). Among them, about 84% has carbonaceous chondritic properties, with the other 16% being associated with ordinary chondrites (Genge 2006; Levison et al. 2009), while basaltic micrometeorites are very rare (Gounelle et al. 2009). However, since micrometeorites can exist as orbiting submillimeter particles only for a few Myr and then, once on the Earth’s surface, are destroyed by the Earth’s geological activity, microxenoliths sample more ancient epochs of the zodiacal cloud than micrometeorites. At the same time, micrometeorites are samples of the zodiacal cloud at 1 AU while microxenoliths, having been embedded in meteorite parent bodies, represent the zodiacal cloud dust that resided in the region of the asteroid Main Belt, i.e., between 2 and 4 AU, or even farther out in the Solar System.
Indeed, objects with sizes up to a few 100 μm, as the micrometeoroids that are the microxenolith precursors, can orbit around the Sun for only a few Myr, because of the Poynting-Robertson and solar wind drag (Leinert et al. 1983; Gustafson 1994). If they enter the Earth’s atmosphere, they can reach the surface and become micrometeorites, but their terrestrial residence time is limited, given the geological activity of the Earth: only rare cosmic spherules—completely melted micrometeorites—show terrestrial ages up to approximately 470 Myr (Meier et al. 2010). In contrast, if they encounter a meteorite parent body, they can be embedded in it and reside in it for very long periods, because of the absence of geological activity. Calculations based on asteroid absolute magnitudes, colors, and ages show that the weathering time scale for S-type asteroids is about 2 Gyr, and that the gardening time scale (i.e., the time needed to refresh an evolved surface) is about 4.4 Gyr (Willman and Jedicke 2011). Given these long time scales of evolution, microxenoliths have the possibility of residing undisturbed on asteroid surfaces for hundreds of millions of years or even more. However, violent impacts, such as those capable of resetting the K-Ar chronometer (Bogard 1995; Swindle et al. 2009), can destroy the regolith layer in which microxenoliths are embedded. The H chondrite impact age distribution shows that such violent impacts happened in two distinct periods (1) before 3.5 Gyr ago and (2) more recently than 1.5 Gyr ago (Bogard 1995; Swindle et al. 2009). If the host meteorites of microxenoliths described in this work come from parent bodies that suffered violent impacts more recently than 1.5 Gyr ago, then their microxenoliths were embedded fairly recently, 1.5 Gyr ago or less. On the contrary, if the host meteorites analyzed herein come from parent bodies for which the K-Ar chronometer was reset only prior to 3.5 Gyr ago, then their microxenoliths could have been embedded in very ancient epochs, as early as 3.5 Gyr ago. The impact age is known only for a few of the H chondrites analyzed in this work (Table 6). A prominent case, however, is Plainview 1917: we found five microxenoliths in this meteorite, and several other xenoliths were previously described (Table 1). For this meteorite, an Ar-Ar impact age of 3.6 Gyr was calculated from the study of impact melt clasts (Keil et al. 1980), supporting the idea that their xenoliths and microxenoliths could have been embedded in very ancient epochs.
Table 6. Characteristics of analyzed H chondrites.
An additional important point is that in the past the flux of submillimeter particles in the inner solar system was higher than in recent times (see e.g., Wasson et al. 1975), so that for a given asteroid the probability of receiving micrometeoroids was higher in more ancient epochs than in recent ones. Therefore, while micrometeorites are samples of the present-day zodiacal cloud dust (up to about 10 Myr ago, given their typical time scale of orbital evolution), microxenoliths could be samples of far more ancient times, up to a few Gyr ago (Briani et al. 2011).
We can hence infer a general feature of the ancient zodiacal cloud dust in the asteroid Main Belt from the properties of xenoliths and microxenoliths observed in this work: their abundance (about 87% of the newly observed xenoliths and microxenoliths) and their mean size (209 μm) indicate that also in ancient epochs the zodiacal cloud complex was dominated by the same submillimeter particles that are the most abundant today (Grün et al. 1985; Nesvorný et al. 2010).
In the following section we describe the relations between newly observed xenoliths and microxenoliths and properties of the host meteorites. In particular, we examine possible correlations with the brecciated nature of the host meteorites and with their petrologic type. Table 6 provides a summary of the analyzed sections, their properties, and number of discovered xenoliths and microxenoliths.
Relation with Brecciated, Gas-Rich Meteorites
An important result of this study is that all the new xenoliths and microxenoliths have been found in brecciated meteorites.
We analyzed samples from 20 brecciated meteorites, 10 of which are regolith breccias. Xenoliths or microxenoliths are present in 6 of these 20 meteorites, 5 of which are regolith breccias. We also examined 20 nonbrecciated meteorites, in which we did not observe any xenoliths and microxenoliths. Of the carbonaceous chondrite xenoliths described in previous works (see Table 1), only one was found in a host meteorite, the H4 Holyoke, that was not a breccia (McCall 1973; Wilkening 1977). In contrast, carbonaceous chondrite xenoliths have been described in eight H chondrite breccias. These data indicate that the probability of finding a carbonaceous chondrite xenolith is clearly larger in a brecciated than in a nonbrecciated H chondrite.
Considering only regolith breccias, i.e., gas-rich meteorites, we have analyzed 10 samples: Dimmitt (H3), Willard (b) (H3.6), Tysnes Island (H4), Weston (H4), Leighton (H5), Plainview 1917 (H5), Pułtusk (H5), Abbott (H3-6), Bremervörde (H/L3.9), and Hainaut (H3-6). In five of these (Willard (b), Tysnes Island, Leighton, Abbott, and Plainview 1917) we found xenoliths. Also, xenoliths have been previously found in Weston (Noonan et al. 1976), Dimmitt, Bremervörde (Goswami et al. 1984), Ipiranga (Rubin and Bottke 2009), and Pułtusk (Wilkening 1977; Goswami et al. 1984). Therefore, gas-rich H chondrites have a high probability of being xenolith-bearing. It is difficult to estimate what percentage of gas-rich meteorites contains xenoliths, because we do not know how many gas-rich meteorites have been surveyed in detail for the presence of xenoliths or microxenoliths. However, combining our results with those present in the literature, we can say that xenoliths and microxenoliths are present in 90% of the regolith breccias in which they have been searched for.
An essential question for the presence of xenoliths and microxenoliths is the size of the parent bodies of the host regolith breccias. To form a regolith layer, fragments produced by an impact must stay on the target body. This implies that a minor body must be sufficiently large for its gravity to prevent the dispersion of fragments produced by the impact of a foreign body. This is clearly the situation observed for howardites, breccias usually considered as coming from the asteroid 4 Vesta, which is the second largest asteroid (with a diameter of approximately 530 km). This consideration of the parent body size is important for the recent scenario proposed by Rubin and Bottke (2009), in which asteroid families, rather than single asteroids, would be the parent bodies of H chondrites. Although asteroid family members (i.e., relatively small asteroids) have the advantage of having low escape velocities, which allow micrometeoroid impacts at low speeds, they also could be too small to develop regolith layers capable of embedding micrometeoroids. A low escape velocity does not favor the formation of a regolith layer, because only the impact ejecta with the smallest velocities can be retained by small targets. However, recent experiments have shown that for a target with limited porosity and low compressive strength (i.e., compact targets) the fraction of impact ejecta with velocities less than 10 km s−1 is quite important: 60% for a 1 mm Al projectile with nominal impact velocity of 4 km s−1 (Michikami et al. 2007). Impact ejecta with velocities less than 10 km s−1 are retained on targets with a radius of 9.5 km and density 2 g cm−3. For a density of 3 g cm−3, a radius of about 8 km is sufficient to retain such ejecta. Asteroids of radius less than 10 km appear to be quite abundant in several Main Belt asteroid families (Zappalà et al. 2002; Durda et al. 2007). These data suggest that asteroid family members with a superficial layer of regolith are not rare, as confirmed also by the fact that all asteroids visited by spacecrafts have well-developed regolith (see e.g., Lee et al. 1996), even the tiny Itokawa, whose maximum dimension is 540 m (Miyamoto et al. 2007; Tsuchiyama et al. 2011). Therefore, asteroid family members can be the parent bodies of H chondrite regolith breccias that contain carbonaceous chondrite xenoliths and microxenoliths.
Considering only regolith breccias, the total sample surface analyzed in this work is 28.5 cm2. The microxenolith number density in H chondrite regolith breccias is therefore 0.7 microxenoliths cm−2. If the entire meteorite surface analyzed in this work is considered (95.2 cm2) then the microxenolith number density is 0.2 microxenoliths cm−2. Howardites show a generally higher content of carbonaceous chondrite xenoliths (Zolensky et al. 1996). In particular, they are very rich in carbonaceous chondrite microxenoliths. Gounelle et al. (2003) observed 71 microxenoliths in only three polished sections of the howardites Yamato-793497, Jodzie, and Kapoeta (all regolith breccias), with a total analyzed surface of 3.8 cm2, i.e., a microxenolith number density of 19 microxenoliths cm−2. These data show that microxenoliths are more abundant in howardites than in H chondrites by two orders of magnitude, and also that their abundance in gas-rich H chondrites is lower by a factor of about 30 with respect to howardites. A possible explanation for the higher abundance of microxenoliths in howardites could be that the orbital properties of Vesta in some way favor embedding of microxenoliths. However, this appears not to be the case: results of dynamical simulations of micrometeoroid orbital evolution (Briani et al. 2011) show that the micrometeoroid flux on Vesta is not larger than on possible H chondrite parent bodies, such as Hebe, Flora, Eunomia, Koronis, and Maria. A second possibility to explain the higher abundance of microxenoliths in howardites than in H chondrites is provided by their impact ages. The distribution of H chondrite impact ages (Swindle et al. 2009) shows several impacts more recent than 1 Gyr. Instead, the distribution of howardite impact ages does not present evidence for impacts more recent than 3 Gyr (Bogard 1995). Therefore, it is possible that the relatively recent impacts suffered by the H chondrite parent body eliminated the previously formed regolith, and the microxenoliths embedded in it, while the accumulation of microxenoliths on Vesta has proceeded quite undisturbed for the last 3 Gyr.
Relation with Host Meteorite Petrologic Type
As reported in Table 6, several meteorites have been analyzed for each petrologic type from 3 to 6. The distribution of xenoliths and microxenoliths as function of the petrologic type is shown in Fig. 8A. From this figure it is evident that most of the new xenoliths and microxenoliths (16 of 24) have been found in H5 chondrites. However, all the microxenoliths that we found in H5 chondrites were found in only two meteorites, namely Leighton and Plainview 1917 (Table 6). H5 chondrites are the most xenolith-rich even considering our results along with those of previous work (Fig. 8B). Most of the xenoliths and microxenoliths previously found in H5 chondrites (four of six) are also concentrated in Plainview 1917. We found no xenoliths or microxenoliths in H6 chondrites. However, we analyzed only H6 chondrites that are not regolith breccias (Table 6). Regolith breccias are the meteorites for which there is the highest probability of finding xenoliths and microxenoliths (as discussed in the previous section). Therefore, the absence of xenoliths and microxenoliths in the H6 chondrites that we analyzed could be due to the fact that they are not regolith breccias. Indeed, we found two xenoliths in Abbott, which is an H3-6 regolith breccia, and we found microxenoliths in Sahara 00181, an H4-6 breccia.
Our results and previous works show that there is no evidence of a correlation between the presence of xenoliths and the petrologic type of their host meteorite. Rather, xenoliths and microxenoliths appear to be common even in relatively high petrologic type H chondrites (H4 and H5). Therefore, we conclude that the embedding of microxenoliths took place after the end of metamorphism on the host meteorite parent bodies, otherwise their fine-grained matrix, rich in hydrous minerals such as phyllosilicates, would not have been preserved (xenoliths and microxenoliths could have been embedded also before or during the parent body thermal metamorphism, but in this case they became so recrystallized that they were rendered very different from samples described herein). Moreover, the presence of xenoliths and microxenoliths in H4 and H5 chondrites indicates that these rocks have been exposed to space after they have been metamorphosed. Therefore, our results require either that the H chondrite parent bodies had a rubble-pile structure after the end of the metamorphism, or that deep, highly metamorphosed layers of a large object were ejected into space to become the H chondrite parent bodies (e.g., an H chondrite parent body totally composed of H5 material).
In the first case, we want to point out that a metamorphosed rubble-pile parent body could have existed early in solar system history. Models for the thermal history of the H chondrite parent body suggest that the different H chondrite petrologic types are due to different burial depths in the parent body (Minster and Allegre 1979; Miyamoto et al. 1982). These onion-shell parent bodies may have cooled undisturbed (Trieloff et al. 2003; Kleine et al. 2008), but the lack of an inverse correlation between metallographic cooling rates and petrologic type (Grimm 1985; Taylor et al. 1987) supports a scenario in which the H chondrite parent bodies were disrupted and reassembled before complete cooling. Our results do not help in discriminating between these two scenarios. However, in both the scenarios (undisturbed cooling and disruption, and reassembly), a rubble-pile parent body could have been present early in solar system history, because (1) undisturbed cooling was completed earlier than 4 Gyr ago (Trieloff et al. 2003; Kleine et al. 2008) and (2) collisions capable of disrupting asteroids were frequent in the early solar system (Grimm 1985) and the reassembly processes were rapid (Grimm et al. 2005).
In the case of metamorphosed layers excavated and ejected into space to become metamorphosed (e.g., H4, H5, or H6) H chondrite parent bodies, such fragments could have formed the present-day asteroid families. As noted by Rubin and Bottke (2009), asteroid families with the right taxonomic type to be H chondrite parent bodies have very ancient ages (i.e., the families were formed a long time ago). The estimated ages of the Eunomia, Koronis, and Maria families are 2.5 ± 0.5, 2.5 ± 1, and 3 ± 1 Gyr, respectively (Nesvorný et al. 2005). These asteroid family ages are also consistent with the previously discussed hypothesis that the H chondrite microxenoliths could have been embedded during very ancient epochs.
In summary, we want to highlight that in both cases (rubble-pile parent body and metamorphosed layers ejected in space to become the H chondrite parent bodies) an H chondrite parent body with surface layers composed of metamorphosed material may have been present earlier than 3 Gyr ago. Also, regolith already existed on the surface of the H chondrite parent body at that time, as attested by the impact age of several impact melt clasts present in regolith breccias (Keil et al. 1980; Swindle et al. 2009). In conclusion, xenoliths and microxenoliths could have been embedded in their host meteorites in very ancient epochs, even earlier than 3 Gyr ago.
In this article we have described our systematic search for new xenoliths (>1 mm in maximum dimension) and microxenoliths (<1 mm in maximum dimension) in H chondrites. Our results show that xenoliths and microxenoliths in H chondrites are not isolated oddities, even if they are not very common. Xenoliths or microxenoliths are present in 10 of 66 analyzed sections, corresponding to 7 of 40 analyzed meteorites. Therefore, in about 17% of the analyzed H chondrites there are xenoliths or microxenoliths.
All the xenoliths and microxenoliths studied herein are composed of carbonaceous chondritic material and about 87% of the newly observed xenoliths and microxenoliths are microxenoliths (<1 mm in maximum dimension). Our results show that carbonaceous chondritic microxenoliths are by far the most common in H chondrites. All the newly observed microxenoliths are dominated by a phyllosilicate-rich, fine-grained matrix. We classified microxenoliths on the basis of their directly observed properties, such as the abundances of olivine, pyroxene, and magnetite. We defined three groups: “magnetite-rich, hydrous,”“magnetite-rich, semi-hydrous,” and “magnetite-poor, semi-hydrous” microxenoliths. Among the 21 new microxenoliths, seven are magnetite-rich, hydrous; ten are magnetite-rich, semihydrous; three are magnetite-poor, semihydrous; and one is ungrouped. The three new xenoliths (size >1 mm) are similar to CM material. They contain abundant chondrules or chondrule fragments several 100 μm in size, set in a phyllosilicate-rich fine-grained matrix composed of serpentine and tochilinite–cronstedtite.
Compositional differences between H chondrite xenoliths and microxenoliths, and the variety of microxenoliths suggest that microxenoliths are not fragments of larger objects impacting onto the H chondrite parent body. Rather, this study supports the idea that microxenoliths are samples of the zodiacal cloud micrometeoroids that populated the inner solar system in ancient epochs (Gounelle et al. 2003). In a companion paper (Briani et al. 2011) we describe dynamical simulations of the micrometeoroid orbital evolution aimed at analyzing the possible origin of H chondrite microxenoliths.
All the newly observed xenoliths and microxenoliths were found in host meteorites that are regolith breccias, i.e., in meteorites formed by compaction of parent body superficial regolith layers. We found xenoliths and microxenoliths in H chondrites with petrologic type varying from 3 to 5. This shows that their embedding happened after host meteorite alterations due to thermal metamorphism. Also, the presence of microxenoliths in H4 and H5 chondrites is consistent with previous observations (Rubin et al. 1983) and models (Grimm 1985; Taylor et al. 1987) suggesting a rubble-pile structure for the H chondrite parent body. These data suggest that the xenoliths’ and microxenoliths’ embedding took place when the host meteorites were still part of the parent body, rather than during their transfer toward the Earth. Combining this with the impact age distribution of H chondrites we conclude that xenoliths and microxenoliths were embedded in the H chondrite parent body early in the solar system history, with possible embedding ages as old as 3.5 Gyr.
A last reflection can be made about xenoliths and microxenoliths that have not been found. In particular, we showed that almost all the newly observed xenoliths and microxenoliths are composed of carbonaceous chondritic material. This means that new types of xenoliths and microxenoliths, radically different from known extraterrestrial materials, and hence with parent bodies possibly different from asteroids, have not been identified. This is the case also for previous studies present in the literature. Such a general result can be explained in two ways: either sources different from asteroids (e.g., comets and trans-neptunian objects) are far less important to produce fragments that later become xenoliths and microxenoliths, or the distinction between solar system minor body populations, asteroids on the one hand, comets on the other hand, is not as sharp as believed. Indeed we support the second hypothesis, as we found that the contributions of asteroidal and cometary micrometeoroids to the H chondrite (and howardite) microxenolith populations are fairly similar (Briani et al. 2011).