The mid sixteenth century Dieppe portolan charts are well-known for their depiction of Java la Grande, a continental land mass, south of the Indonesian archipelago. Its identity is a persistent problem of early mapping. According to some opinions it is a depiction of Australian coasts, according to others it is an imaginary land made up of coastlines that either belong elsewhere or have been drawn from imagination. Even more problematic are the three groups of islands depicted in the south-west Pacific Ocean on some of the same charts. It has most often been assumed that they are geographically related to Java la Grande so that if, for example, the land mass is supposed to depict Australia then they must be depictions of islands in the south-west Pacific; if its eastern coastline is supposed to be the coast of Indo-China then they must be islands in the South China Sea; if it is supposed to be a duplicated South America then they must be Caribbean islands. The evidence for such identifications is weak and the arguments are unsatisfactory.

There is sufficient internal evidence to indicate the origins of these island groups. The evidence comprises spatial information (notably latitudes, distances and bearings, which can be obtained by cartometry), the nomenclature, and the configurations of the islands. There is also external evidence in the forms of contemporary texts on the geography of the South Pacific and of non-Dieppe mapping.

In this paper it will be argued that the three groups of islands originate in contemporary accounts of Magellan’s 1519-1522 expedition. In 1521 Magellan found two small, barren, uninhabited and harbourless islands in the middle of the South Pacific. It will be seen that the three island groups of the Dieppe maps are derived from the inconsistent treatment of the two islands in contemporary accounts, on one hand in the account of Maximilianus Transylvanus, and on the other hand in several eye-witness accounts including Antonio Pigafetta. The development of the two ‘unfortunate’ islands into three large and attractive groups can be traced partly in the texts of Jean Alfonse (also called Jean Fonteneau de Saintonge), partly in the nomenclature, and partly in their shapes on the Dieppe maps. The origins of the groups of islands can barely be seen behind the Dieppe cartographers’ cosmographical speculations and remodelling.