Did the first insects live in water or in air?

We evaluate the arguments and evidence for a terrestrial vs an aquatic origin for the Insecta. The evidence falls into three categories: ( I ) evidence that does not support one view more than the other; (2) speculative evidence, which should carry little weight; and (3) evidence that does support one view more than the other. Category I includes evidence from locomotory and osmoregulatory systems; plausible functions have been proposed for 'protowings' in both aquatic and terrestrial environments, while locomotory and osmoregulatory mechanisms of insects shed little light on their origins. Fossils, phylogenetic speculation, gill structure and life histories fall into category 2, in which, although speculative, the evidence favours a terrestrial origin. The earliest fossil hexapods were apparently terrestrial and unequivocally aquatic hexapods do not appear until 60-70 million yean later, while sister-group relationships point to a terrestrial life style from at least the Hexapoda-Myriapoda stem group. The great variation in gill structure, even within orders, suggests convergence, and the more or less completely aquatic life histories are better interpreted as steps towards independence from land, rather than signs of an aquatic origin. Category 3 includes evidence from the tracheal system. In order to have evolved in water, a tracheal system must have first invaginated, then connected with the body wall for gas exchange with the water, and thirdly connected with the internal organs. I t is difficult to envisage functions for the first two stages; on the other hand, the system could have readily evolved on land by invagination of respiratory surfaces, and then have been modified to effect gas exchange in water via gills.