Komatiites are igneous rocks mainly of Archaean age. Their ultramafic origin is controversial and has been attributed to either high degrees of dry melting of peridotite in deep mantle plumes at temperatures up to 500°C hotter than today, or to melting at shallow depth, in volatile-rich (hydrous) mantle with temperatures not much different than the 1200–1500°C (for basalts and boninites) in contemporary mantle of subduction zones or below Archaean oceanic ridges. Komatiites often have exquisite disequilibrium mineral textures seldom seen in younger rocks. These spinifex textures give komatiite a mystical aura, keenly sought after, because a detailed understanding of their origin remains shrouded in the dissipated temperature and volatile gradients of komatiite magma. These unresolved issues thwart fundamental understanding of how the young, early Earth worked. Komatiite explores these issues in some depth. The principal author, Nick Arndt, has studied these rocks for nearly 40 years, from just after their discovery in South Africa, and has observed as many komatiites as anyone. His co-authors, together, have clocked up more than 50 years of experience on ore deposits associated with komatiites. Their book describes komatiites and their geological settings in considerable detail; all komatiites that have ever been found and described to date appear to be included. This well-illustrated book comprises two parts, both subdivided into seven chapters, and comes with an extensive bibliography of more than 700 references, a useful index, and four colour plates. Part I classifies and describes various aspects of komatiites: textures, field characteristics, mineralogy, geochemistry, isotope systematics and experimental data. Part II provides interpretations concerning the volcanology, origin, emplacement, economic value, tectonic setting and geodynamic evolution of komatiite magma.

Nick Arndt's previous (co-edited) book on these rocks, in Reference Arndt and Nisbet1982, was called Komatiites. A change to the singular in the title of this new book is symptomatic of present global petrologic group-think: Nick Arndt, his co-authors, and co-workers, are unapologetic followers of dry, white-hot, fluid komatiite lava derived from deep mantle plumes – one model fits all. But in reality komatiite has a very large range of composition (18–30% MgO, 44–52% SiO₂, etc). If they were modern rocks they would have five to ten different names. Clearly there are komatiites and komatiites, and it is perhaps unlikely that one idea (wet or dry) or tectonic setting will explain them all. That is like lumping basalts, boninites and andesites together and then asking whether they form wet or dry in one tectonic environment. Dry komatiites occur in some instances like, for example, the late Archaean Zimbabwe continental flood komatiites extruded subaqueously across and/or marginal to their craton as they do along present passive continental margins (e.g. Hynes, Reference Hynes, Condie and Pease2008). However, the observation that some minerals (pyroxenes) in the original type locality of komatiites (Barberton) can only be reproduced by hydrous experiments still stands. No-one has come up with a viable alternative explanation, despite attempts in this present book; one might as well take the arguments to the Moon where once one might have attempted to argue that such pyroxenes (augites) be used against wet komatiites, but now (since this book was written) new evidence suggests that even there the mantle of the early Moon might have been wet (Saal et al. Reference Saal, Hauri, Lo Cascio, Van Orman, Rutherford and Cooper2008). Arndt argues that wet magmas would not retain their volatiles before reaching the surface. But there are modern, demonstrably hydrous, mafic magmas (boninites) that erupt as lava flows on the sea floor and preserve high water contents (upwards of 5%) without degassing or exploding. Barberton and other South African komatiites (e.g. Commondale) are similar to boninitic lava flows produced from wet magmas. Arndt thinks this is impossible, and to him proving that komatiites are lava flows equates to proving that they are anhydrous. Even if this were true, the field evidence that all komatiites are lava flows is equivocal, including in the type area of Barberton, despite Arndt's certification that recent mapping there has shown this beyond doubt, ‘destroying the very foundations of the hydrous komatiite model’. There are plenty intrusive examples there; and in Australia and Zimbabawe too, where occasional cross-cutting relationships with host rocks have been interpreted as thermodynamic flow-erosion by some researchers and as intrusive contacts by others. The real issue here, which is not addressed at all in this book, is how these field relationships might be resolved to facilitate better reconstruction of the komatiite-magma plumbing systems that will help better understand the construction of Archaean lithosphere. There is a tendency also to apply inconsistent chemical criteria to these ubiquitously (and pervasively) altered rocks – as are all komatiites. For example, komatiites generally have 0.5–1 wt% TiO₂, whilst boninites have less than 0.5 % (by definition). There is some overlap, but they are distinct. Komatiites have slightly higher TiO₂ than boninites (except for Commondale that really looks like a boninite). Plots of various ratios with TiO₂ are used to argue that this 0.5 wt% difference proves that boninites and komatiites are completely different and that komatiites are plume magmas. Yet all assumed plume magmas have over 1 wt% TiO₂ and generally more than 2 wt% and there is no overlap with the komatiite data at all. The chemical traces left by Archaean processes need much more robust and careful detective work.

In the end, however, my real gripe with views expressed in the book lies in the often autocratic manner of self-fulfilling arguments. Interpretations are seldom as clearly separated from the observations and descriptions as the division into Parts I and II imply, and subjectivity is present throughout. Therefore, the book is an editorial contribution to the science of komatiites. That's OK – Arndt and his co-authors have put enough time in these rocks to deserve their say, but they cast aside the results of carefully conducted hydrous experiments too easily. That said, they have done a good job in collating the existing data. Komatiite is a good ‘Wikipedia’ start for graduate students interested in the early Earth, and petrology in general, provided some guidance is at hand.