19.10 N&V mx

but only rarely is a new type of meteorite discovered. Last week in Science, Brown et al. described what might be such an event with the orbit, fall and recovery of numerous fragments of a meteorite in the Canadian Yukon. Named after the frozen lake on which it fell, the Tagish Lake meteorite is a chondrite, which means it is a stony aggregate of matter from the early Solar System. It is described as a carbonaceous chondrite — one which contains relatively high levels of carbon — but it also has unusual properties that make it difficult to classify properly. All chondrites are samples of matter from the early Solar System that have undergone only minor geological processing. As such they can provide us with a snapshot of some of the earliest material in the Solar System. But the chemical composition of the Tagish Lake meteorite suggests that it might be especially primitive. In other words, its composition is very similar to that of the Sun. Much of what we know about the abundance of elements in the Sun is derived from analysis of carbonaceous chondrites, some of which bear more than a passing resemblance to a dried-out mud puddle. The luminous centrepiece of our Solar System would appear to have little in common with a lump of mud. But spectroscopic estimates of elemental abundances in the Sun’s visible surface, known as the solar photosphere, match most closely with the compositions of these meteorites. Consequently, trace elements in the Sun that cannot be measured spectroscopically are determined using elemental abundances in chondrites. Despite their primitive chemistry, carbonaceous chondrites typically contain clays, oxides and other minerals, sometimes cemented by veins of carbonate and sulphate. Mineralogically, they are more ‘cosmuck’ than cosmic. The mineral alteration is thought to occur within the parent asteroids, when accreted ices are melted by heat produced during the decay of shortlived radioactive nuclides. Like other carbonaceous chondrites, the Tagish Lake meteorite has reacted with water at low temperatures, but its mineralogy appears to be somewhat less altered (see Box 1). Brown and colleagues speculate that the Tagish Lake meteorite might be a sample of the material that was the precursor of the CI chondrites, the group of chondrites until now thought to be the most chemically primitive group. As such, they say, it may be a new type of carbonaceous chondrite. The relative proportions of oxygen isotopes vary considerably among chondrites and are commonly used in meteorite classification. The oxygen-isotope composition of Tagish Lake places it among the CI chondrites (see Box 1). But in terms of texture, the meteorite resembles another carbonaceous chondrite group, the CM chondrites. As a result, this meteorite may force meteoriticists to revise their classification system. It is the chemistry of the Tagish Lake meteorite, rather than its mineralogy, that is probably most intriguing. Samples of the still-frozen meteorite were collected soon after it fell. As a result, they have probably suffered minimal terrestrial contamination. In contrast, museum samples of many CI and CM chondrites have been stored under less than sterile conditions. Brown and coworkers have analysed the abundances of most elements in the Tagish Lake material, and their analyses offer an interesting comparison with the Sun’s composition. Meteoriticists divide elements according to their volatility, because elements are partitioned among and within chondrites based on the temperatures at which they condensed. The abundances of the least volatile elements in Tagish Lake resemble those in CM chondrites, whereas the more highly volatile elements lie between CM and CI values. Comparing elemental abundances, Brown and co-workers found that CI chondrites provide a closer match to solar composition for 39 analysed elements, whereas Tagish Lake provides a closer match for just 17 elements. From this comparison, it is not obvious whether Tagish Lake is more primitive than CI chondrites, as the most primitive sample would be expected to match the