Sir James was born on 6 September 1901 at Cambridge where he attended the Perse School. He received his geological education at Chelsea Polytechnic and the Royal College of Science. In 1923 he was appointed demonstrator in geology at the Imperial College of Science and Technology, London University, where he stayed until 1928; for his last two years there he also served as Warden at the pioneer Imperial College hostel. While at Imperial College he studied the Shineton Shales of the Wrekin district of Shropshire under the guidance of W. W. Watts and in collaboration with his life-long friend O. M. B. Bulman, and was awarded a Ph.D. in 1925. This work in Shropshire provided the stimulus for his continuing contribution on the Cambrian system and its faunas, and on trilobites of all ages. During this period he also recorded his observations on Tertiary crabs from Zanzibar, a group he was to return to in later years. In 1928 he obtained one of the two posts offered by the Geological Survey, being appointed Geologist. He undertook field work in the Dorking district as a prelude to an intended posting to the Survey office at York, but the death of G. W. Lee, then palaeontologist in Edinburgh, determined that Stubblefield should remain at the headquarters of the Survey at Jermyn Street in London. Internal transfers of the palaeontological staff left no member available to determine the fossils then accumulating from the current survey of the Shrewsbury district, and Stubblefield was asked by the Director to undertake this task because of his knowledge of the Lower Palaeozoic faunas of Shropshire. This transfer became long-term and thus began his association with, and eventual leadership of, the Palaeontological Department of the Survey, and secured the continuation of his notable contributions to palaeontology. The Shrewsbury commitment led to visits to the area during which new faunal horizons were discovered in the local Cambrian and Ordovician, including the Nemagraptus gracilis fauna from the Breidden Hills; other finds included the then earliest British eurypterid subsequently described by L. Størmer as the type of a new genus under the name of Brachyopterus stubblefieldi.

Cretaceous-Palaeogene sediments of the Kangerdlugssuaq area on the continental margin of Central East Greenland were deposited in an embayment of an extensive pre-NE Atlantic shelf sea. Pre-Sparnacian sediments are thin (150 m), incomplete and of siliciclastic type, formed in shallow marine waters. Sparnacian times saw the onset of vigorous basaltic vulcanicity, marking the initial rifting episode between Greenland and Eurasia. Uplift immediately prior to the vulcanicity is evidenced by an unconformity at the base of the Sparnacian, above which basement-derived arkosic sandstones and conglomerates are followed by about 1.5 km of coarse volcaniclastics, basaltic flows of dominantly picritic composition, pro-grading hyaloclastite wedges and thin siltstones with abundant organic detritus. Very rapid subsidence accompanied this early phase of vulcanicity, maintaining the top of the pile close to sea level and allowing the deposition of a further kilometre of waterlain tuffs in the embayment. Sedimentation extended northwards and eastwards on to basement rocks at this period, with the formation of a non-marine sequence which includes coals.

The overlying plateau tholeiites overlap the earlier volcanics; their depositional area was extremely extensive along the East Greenland margin and bears no relationship to the Kangerdlugssuaq sedimentary embayment, although their thickest development, 4 km or more, was attained in that region. Eruption rate of the pile exceeded subsidence for a period and it is dominantly subaerial.

This sequence of events is compared with the similar history of sedimentation and basaltic vulcanicity on the west coast of Greenland, and it is inferred that just as the East Greenland sequence records the initiation of spreading between Greenland and Rockall-Faeroe at anomaly 24−25 time, so that of West Greenland marks the propagation of the Labrador Sea spreading axis through the Davis Strait into Baffin Bay at anomaly 26−27 time.

Seatura is a Pliocene caldera-type, shield volcano situated in SW Vanua Levu, Fiji. Shield-building basaltic lavas are predominantly subaerial though there is a shoreline facies of fossiliferous sediment, pillow lava and hyaloclastite. Late-stage events include the intrusion of a dyke swarm in and around the caldera, and deposition of laharic conglomerates. Chemically the rocks are intermediate between tholeiite and alkali basalt and have intraoceanic levels of TiO2. Late-stage derivatives include benmoreite, trachyte, and propylitically altered rocks. The parent magmas are thought to have been partial melts of undepleted mantle at relatively deep levels. Tapping of these deep levels may have been facilitated by extensional plate tectonic conditions existing in the Fiji area during the development of Seatura Volcano.

The Devonian System of China occurs in the Tianshan–Hingan* Geosyncline of North China, the Yangtze Paraplatform of South China and in the geosynclinal areas of West China. Devonian formations are entirely absent from the Sino-Korean Paraplatform and the Tarim Platform, but along the border of the Tarim Basin and the slopes of Qilianshan (Chilienshan) are continental sequences rich in vertebrate and plant fossils. On the basis of differences in biostratigraphic characteristics the Devonian is divisible into 8 regions: 1. Junggar–Hingan (Dzungar–Khingan), 2. South Tianshan (South Tienshan), 3. Qilianshan (Chilienshan), 4. Longmenshan (Lungmenshan), 5. South China, 6. South-eastern China, 7. West Sichuan–North Xizang (West Szechuan-North Tibet), 8. Himalaya–West Yunnan.

Using both lithology and fossil content, five principal facies (types) may be further differentiated: 1. Hingan type (eugeosynclinal facies), 2. Baoxin type (miogeosynclinal facies), 3. Xiangzhou type (platform near-shore facies), 4. Nandan type (platformal off-shore facies), 5. Cuifengshan (Chuifeng shan) type (continental facies).

The Devonian of South China occurs in various facies, but has a well-defined principal stratotype. It has been divided by Chinese geologists into eight stages, two in the Upper Devonian, two in the Middle Devonian and four in the Lower Devonian.

The Devonian sequences of the geosynclinal regions of West China are characterized by facies which contain graptolites and tentaculites, and have largely been affected by regional metamorphism. The Devonian biota of the Junggar–Hingan Region is essentially an endemic one, but it is associated with a few European and North American elements. Marine pyroclastics are the principal rocks. General stratigraphical columns of those regions are given with a correlation with other parts of China. The continental Devonian of China is unique especially in the light of its vertebrate content. Red sandstones are widespread in the Lower Devonian, and intercalated with argillaceous limestone and mudstone. This sequence is rich in endemic East Asiatic faunal elements including agnatha (Galeaspida, Polybranchiaspida), Antiarchi, etc. Middle and Upper Devonian lithostratigraphic units consist principally of light coloured quartzose sandstone formations in Southeastern China, with some red beds in South and North China, while their fauna and flora are cosmopolitan in character. Their exact geological ages vary, however, as clearly shown in the South China region, especially towards the top of the Upper Devonian (with Asterolepis) and in the early Middle Devonian (with Bothriolepis).

Despite their importance and frequency of usage, ‘grouping terms’ in petrology such as Association, Family, Lineage, Province, Series and Suite appear to lack established meanings. Partly by analogy with the recently standardized stratigraphic nomenclature, hierarchies of such terms are suggested to cover petrographic classification (rock-names), petrological classification (rock-names extended to magma-types), geographic classification (igneous provinces) and tectono-genetic classification (rock assemblages characteristic of particular tectonic regimes). Used in this way, the exact implications of each term become clearer. Again by analogy with stratigraphy, ‘Super-’ and ‘Sub-’ groupings are suggested in a few cases: ‘Subprovince’ for example can differentiate distinct rock assemblages related to separate parent magmas within one well-defined Province (e.g. nephelinite-carbonatite and basalt-trachyte assemblages in the East African Rift), and ‘Superprovince’ can be used to group coeval Provinces now separated by opened oceans, etc. (e.g. Thulean Superprovince composed of Hebridean, East Greenland, etc., Provinces). The term ‘Province’ itself carries several ambiguities; it is best confined to cases where the ages of the spatially associated igneous rocks fall within a narrow range (e.g. Oslo Province); ‘Belt’ is suggested for those where the age-range is more substantial (e.g. Jurassic-Tertiary magmatism of Southeast Brazil).

Llandovery strata at Llanystumdwy are described. Ninety graptolite species have been identified enabling the establishment of the following zones: acinaces, cyphus, triangulatus, magnus, leptotheca, convolutus, ?sedgwickii, turriculatus and crispus. The name Llanystumdwy Mudstone Formation is proposed for previously undescribed mudstones cropping out on the River Dwyfor at Llanystumdwy and includes all strata referable to the acinaces to turriculatus zones. Comparison is made between Llanystumdwy and other areas, notably Northern England and Southern Scotland

Thermal stresses due to contraction during cooling have been widely accepted as the origin of columnar structures in certain igneous rocks. It is proposed that these highly ordered structures could be the result of a double-diffusive convective process in the melt during cooling, a process analogous to salt fingers observed in the laboratory and the oceans. Salt fingers are driven by appropriate differences in heat and salt content at the top and bottom of a stably stratified water column. In molten basalts, temperature and chemical-compositional differences between the top and bottom of a stagnant melt could drive highly regular columnar ‘finger’ motions in the melt. When solidification takes place eventually, cracks due to contraction during cooling would then have preferred paths of propagation into the interior from the surface along the junctions of adjacent ‘basalt fingers’ giving rise to columnar jointing. Morphological evidence supporting the finger hypothesis is cited and useful ways of gathering further field data that would shed some light on the validity of the hypothesis are suggested. For the benefit of geologists not familiar with double-diffusive processes, an up-to-date review of these flow phenomena is included.

The current belief in a single chalky till glaciation of eastern England is upheld by a detailed examination of a section at Barrington which had been said to reveal two such tills. An attempt is made to arrange the glacial, periglacical and interglacial beds of southern Cambridgeshire into stratigraphic order and to place them within Quaternary stages. The study concludes with a consideration of the fabric data that were used to support the claim that two major chalk-rich tills occur in East Anglia. Till mineralization is reported in the area for the first time.

Prior to Late Triassic–Early Jurassic times, the geological history of the Antarctic Peninsula region was dominated almost entirely by Gondwana sequences that together comprised a major arc-trench system. Subduction complex, trench-slope-break and fore-arc basin sedimentation can all be recognized, and deposition was at least partly on early Palaeozoic or older continental crust. The only evidence for a contemporaneous magmatic arc situated in the Antarctic Peninsula at this time consists of patchy occurrences of metavolcanic rocks, possibly representing the frontal edge of the arc, and the major outcrop area of these rocks is believed to lie under the broad shallow continental shelf E of the Antarctic Peninsula. This is contrary to most current hypotheses in which a marginal basin, presumably floored by oceanic crust, is thought to crop out close to the E coast of the Antarctic Peninsula. However, the complete absence of substantial outcrops of pre-Jurassic volcanic rocks anywhere in eastern Antarctica and South Africa, which are the closest and most likely places in which these should exist, supports the new proposal. Moreover, the identification of back-arc elements of the arc-trench system (foreland fold-thrust belt and retro-arc foreland basin) in eastern Antarctica and South Africa greatly strengthens the likelihood of the arc cropping out in the area suggested.

Towards the end of the Triassic and during the Early Jurassic Periods, an intense diastrophic event, or culmination of events, of orogenic magnitude occurred (Gondwanian orogeny), causing substantial redistribution of the pre-existing elements of the arc-trench system. In particular, all the fore-arc sequences were strongly deformed, some possibly for the first time (e.g. in the fore-arc basin), and became firmly accreted to the continental margin. Moreover, the magmatic foci migrated trenchwards to intrude the deformed rocks in the Antarctic Peninsula. Because many of the plutons were emplaced synkinematically, they often closely resemble ‘basement’ gneisses. The distinction between these rocks remains a serious problem in the Antarctic Peninsula and it can be solved only partially by radiometric dating.

Late Triassic Radiolaria have been recovered from a bedded chert collected early in 1977 in the South Orkney islands. This is the first reliable palaeontological evidence for the age of the so-called basement complex that forms most of the exposed rock of a microcontinent on the southern limb of the Scotia Ridge. The fossils place a lower limit on the age of deformation of at least part of the complex as post-Karnian or Norian. The basement is overlain by undeformed Jurassic to Cretaceous conglomerates, which suggests this event took place during the early Mesozoic Gondwanian orogeny.

The new data are consistent with the interpretation that the basement complex of the South Orkney Islands represents part of a subduction complex and was accreted to the Pacific margin of Gondwanaland during the late Triassic or early Jurassic time.

Stubblefield suggested that proparian trilobites may have arisen by paedomorphosis. Examples of the role of paedomorphosis in the evolution of species and of particular organs are reviewed. The cryptogenetic renewal of trilobites after the Cambrian resulted in distinctive major taxa, but their relationships to each other and to older groups remain unsolved. Similarities between early developmental stages should be re-assessed as a possible guide to relationships. The early Upper Cambrian, type species of the proparian Schmalenseeia is re-described. It was small, possibly world-wide in distribution, the exoskeleton thin, the parts not articulated, apparently forming a more or less rigid shield. It may have been a paedomorphic, planktic species, but its origin and relationships are obscure.

K2O, TiO2, P205, Zr, Y, Sr and Nb data from the Panjal Traps of Kashmir indicate that the lava erupted in a rift tectonic setting. This is also suggested by the other features of these rocks and volcanic history of the area. The time of eruption represents the initial phase of break-up of the Indian plate from the Siberian plate. This conclusion is in good agreement with the current views on the tectonic evolution of the area.

Cyclopyge genatenta Stubblefield, 1927, is interpreted from new material, and established as the type species of a new cyclopygid genus, Prospectatrix. The morphology of Prospectatrix shows that the cyclopygids and nileids are closely related, and together with the Taihungshaniidae, these families are included in an expanded concept of the Cyclopygacea. The subfamily Ellipsotaphrinae Kobayashi and Hamada, 1971 is excluded from the Cyclopygidae. Pelagic habits arose on at least five separate evolutionary lines among Ordovician trilobites: Prospectatrix lies near the origin of one of these. Cremastoglottos Whittard was derived from a robergioid ancestor by forward migration of the eyes.

Calcareous nannofossils from mid-Cretaceous (Albian-Cenomanian) sections at Pont du Fahs (northern Tunisia) allow the recognition of three nannofossil zones-the Prediscosphaera cretacea Zone, the Hayesites albiensis Zone, and the Eiffellithus turriseffeli Zone. A comparative study has been made of the nannofossil assemblages from Pont du Fahs and 14 other sites in North Africa, Western Europe, the North Atlantic and United States.

Pelitic schists from the Appin Group of the Dalradian Supergroup are all chemically distinct from pelitic schists in the older Moinian Supergroup. In addition, trace-element geochemistry can also discriminate between the individual pelitic schists within the Appin Group, especially those within the Lochaber Subgroup. These differences apparently relate mainly to differences in provenance rather than differences of sedimentary environment, and are sufficiently consistent to offer a means for correlation or identification of pelitic units in areas of the Central Highlands where the stratigraphy is still in dispute.

In the evolution of the Scottish Dalradian basin, the climax of widespread tholeiitic magmatism pre-dated or coincided with major basement faulting and the onset of deep marine sedimentation in Lower Cambrian times. Earlier basin-deepening and precursory igneous activity had already occurred. Major and trace element chemistry of low-grade metabasites indicates their affinity with basalts from ocean-spreading centres, or transition to ‘within-plate’ types. Geological evidence favours a tectonic setting for this igneous suite and the Highland Border Series on the NW margin of an actively evolving Lower Cambrian ocean basin, possibly a marginal basin to the Iapetus Ocean System.

Eocambrian volcanic breccias of angular dacite and andesite clasts at Wadi Fatima, Kingdom of Saudi Arabia, contain perfectly preserved gas pipe structures. Only the latter identify them as lithic-rich, pyroclastic flow breccias of Pelean type. Without the gas pipes it would not be possible to identify them as pyroclastic and different from epiclastic breccias originating as mudflows, lahars or avalanches. The pyroclastic flow breccias constitute a minor component of the volcanic Fatima Group dominated by pumiceous water deposited sediments and subaerial welded tuffs.

Preserved in the deformed peridotites of the Lizard, Cornwall, are relics of intricate primary precipitates of olivine. The mineral compositions of those rocks with an abundance of primary relics are typical of the primary assemblage Lizard peridotite. In detail the olivine primary structures resemble, but are not identical to, the crescumulates described from various layered ultrabasic intrusions. The available evidence shows distinctive specific features in the Lizard crescumulates in the scale and morphology of the structures and in the conditions of growth.