We have followed what has been written in this Journal about the accuracy of dead reckoning in the air and the determination and use of the most probable position. We wish to refer in particular to the article by J. B. Parker entitled ‘The navigational implications of Mr. Durst's paper’ (this Journal, 8, 113). We are not familiar with the routes of the North Atlantic or Central Africa; we have, however, eight years experience on the routes Lisbon–Luanda–Lourenço Marques, via Dakar, as well as those overflying the Sahara Desert via Oujda and via Agadir. The following considerations are in great part the result of that experience.