Explore the vast range of titles available.

Narratives of the Claudian invasion of Britain in a.d. 43 have regularly referred to elephants being part of Claudius’ force, with some accounts even suggesting that Claudius paraded the beasts through Colchester (Camulodunum), or even rode on top of one. This study investigates these claims, which derive solely from a somewhat ambiguous reference in Cassius Dio's (60.21.2) description of the invasion. Temporal and logistical constraints, together with military and iconographic considerations, however, make it highly unlikely that the animals, even if they had been assembled on the Channel, made their way across to Britain. Overall, the study shows that Dio's testimony should be treated with extreme caution, and should be accorded only parenthetical importance in treatments of the Claudian invasion.

A Roman villa building at Mud Hole, Boxford, West Berkshire, was examined by excavation in 2017 and 2019, and found to be of probable fourth-century date. One room of this otherwise seemingly modest villa contained a remarkable late fourth-century figured mosaic, which features a number of rare mythological subjects not previously encountered in Britain. Inscriptions suggest the name of the villa owner (Caepio) and his wife (Fortunata), with a possible Spanish connection. The mosaic's central panel is ornamented with the triumphs of Pelops and Bellerophon, the former known only from two other mosaics, in Syria and Spain. The borders also contain depictions of stories unknown on other mosaics, but all concerned with aspects of triumph. The central panel is upheld by walking telamones (giants), otherwise only known on a mosaic from Tusculum, and the mosaicists have attempted to use foreshortening to give the floor a trompe l'oeil effect. The rare subjects depicted on the floor all relate to either Poseidon, Pelops, Bellerophon or Atlas, and suggest high standards of mythological knowledge and longevity of classical culture amongst the villa-owning inhabitants of late fourth-century Berkshire. The mosaic shows a connection to earlier depictions of the Pelops story, but is highly original in its interpretation of them and follows a contemporary trend, not previously encountered in Britain, of its subjects breaking out from their ornamental borders. The mosaic is an altogether exceptional discovery and can be considered an important example of late Roman art so far found in Britain.

Lying at the heart of the city of Venta Silurum, the forum-basilica at Caerwent tells the public, municipal, story of the Civitas Silurum in south Wales. The re-excavation of this complex between 1987 and 1995 revealed a wealth of new information that has a great deal to tell us about the history of the city and the fortunes of its inhabitants. Constructed in the early second century and rebuilt in the middle of the fourth century, the forum-basilica was where the administrative, judicial, commercial and religious functions of the city were located. The modifications to the basilica at the very end of the Roman period indicate continued use of this civic space in the later fourth and fifth centuries.

A substantial database of published excavation and other reports has been used to map the character and distribution in Roman Britain of whetstones, those unprepossessing implements essential in the home, farmstead, workshop and barracks for the maintenance of edge-tools and weapons. The quality of the geological identifications in the reports varies considerably, but a wide range of lithologies are reported as put to use: granite, basalts-dolerites, lava, tuff, mica-schist, slates/phyllites, Brownstones, Pennant sandstone, micaceous sandstones, grey sandstones/siltstones, Millstone Grit, Coal Measures, red sandstones, ferruginous sandstones, sarsen, Weald Clay Formation sandstones, sandy limestones, shelly limestones, cementstones, and (Lower) Carboniferous Limestone. On distributional evidence, some of these categories are aliases for alternatively and more familiarly named lithologies. Bringing ‘high-end’ products to the market, the long-running industry based on sandstones from the Weald Clay Formation (Lower Cretaceous) emerges as a British economic feature, evidenced from the Channel coast to the Scottish Borders, and with a recently demonstrated, substantial representation on the Roman near-continent. The distribution maps point to another and more complete British industry, based on the Brownstones (Old Red Sandstone, Devonian) and Pennant sandstone (Upper Carboniferous), outcropping close together in the West Country. A more systematic and geology-based treatment of excavated whetstones in the future is likely to yield yet more insights into the role these artefacts played in the economy of Roman Britain.

The retention of 2.6 metric tonnes of building material from three rural masonry buildings from Bottisham, south-east Cambridgeshire, provided a rare opportunity for a thorough investigation into their fabric, form, construction style and function. A double-apsidal building may have been a bath-house and another building had evidence for an extensive box-flue tile heating system. Both buildings showed signs of either being unfinished or the heating element having never been used. A third building was a later construction that used rare red-slipped tegulae and imbrices. This article goes beyond suggesting the existence of a villa or villa-type complex at Bottisham to offer a detailed case study of the use of ceramic building materials

All bacterial isolates from 7058 patients admitted to haemato-oncology wards at National Taiwan University Hospital between 2002 and 2006 were characterized. In total 1307 non-duplicate bloodstream isolates were made from all patients with haematological malignancy; 853 (65%) of these were from neutropenic patients. Gram-negative bacteria predominated (60%) in neutropenic isolates with Escherichia coli (12%), Klebsiella pneumoniae (10%), Acinetobacter calcoaceticus-baumannii complex (6%), and Stenotrophomonas maltophilia (6%) the most frequent. Coagulase-negative staphylococci (19%) and Staphylococcus aureus (4%) were the most common Gram-positive pathogens. Resistance to ciprofloxacin was found in 50% of E. coli and 20% of K. pneumoniae isolates from neutropenic patients. Extensively drug-resistant A. calcoaceticus-baumannii complex and vancomycin-resistant enterococci were also found during the study period. Emerging antimicrobial resistant pathogens are an increasing threat to neutropenic cancer patients.

We determined the frequency of isolation of Leptospira from dogs and rodents, the serovars of Leptospira, and the clinical, gross and histological manifestations in dogs with leptospirosis in Trinidad. From dogs, samples of urine, blood and kidney were collected while only kidney and blood samples of trapped rodents were used. Isolates were cultured and serotyped using a panel of 23 international serovars and monoclonal antibodies. The risk factors for leptospirosis were also determined in owned dogs using a standard questionnaire. Of a total of 468 animals investigated for Leptospira, 70 (15·0%) were positive, comprising nine (18·0%) of 50 suspected canine leptospirosis cases, seven (3·4%) of 207 stray dogs and 54 (25·6%) of 211 rodents. The observation that rodents have a statistically (P<0·05, χ2) higher frequency of isolation emphasizes the importance of rodents as reservoirs of leptospirosis in the country. Copenhageni was the predominant serovar found in 100·0% (7/7), 33·3% (2/6) and 68·5% (37/54) of isolates from suspected canine leptospirosis cases, stray dogs and rodents, respectively. Serovars Icterohaemorrhagiae and Canicola, the two serovars present in the commercial vaccines used locally, were detected in one (1·5%) and zero (0·0%) isolates respectively of the 67 tested. Data provided suggest that the apparent vaccine failure may be a consequence of the fact that the predominant serovar (Copenhageni) detected in sick, apparently healthy dogs and in rodents is not contained in the vaccines used locally to protect dogs against canine leptospirosis.

The data torrent unleashed by current and upcoming astronomical surveys demands scalable analysis methods. Many machine learning approaches scale well, but separating the instrument measurement from the physical effects of interest, dealing with variable errors, and deriving parameter uncertainties is often an afterthought. Classic forward-folding analyses with Markov chain Monte Carlo or nested sampling enable parameter estimation and model comparison, even for complex and slow-to-evaluate physical models. However, these approaches require independent runs for each data set, implying an unfeasible number of model evaluations in the Big Data regime. Here I present a new algorithm, collaborative nested sampling, for deriving parameter probability distributions for each observation. Importantly, the number of physical model evaluations scales sub-linearly with the number of data sets, and no assumptions about homogeneous errors, Gaussianity, the form of the model, or heterogeneity/completeness of the observations need to be made. Collaborative nested sampling has immediate applications in speeding up analyses of large surveys, integral-field-unit observations, and Monte Carlo simulations.

The ability to build a model on a source task and subsequently adapt this model to a new target task is a pervasive need in many astronomical applications. The problem is generally known in the machine learning field as transfer learning, where domain adaptation is a popular scenario. An example is to build a predictive model on spectroscopic data to identify Type Ia supernovae (SNe Ia), while subsequently trying to adapt such a model to photometric data. In this paper we propose a new general approach to domain adaptation which does not rely on the proximity of source and target distributions. Instead we simply assume a strong similarity in model complexity across domains, and use active learning to mitigate the dependence on source examples. Our work leads to a new formulation for the likelihood as a function of empirical error using a theoretical learning bound; the result is a novel mapping from generalization error to a likelihood estimation. Results using two real astronomical problems, SN Ia classification and identification of Mars landforms, show two main advantages of our approach: increased performance accuracy and substantial savings in computational cost.