In his Presidential Address (Journal 16, 1) Dr. G. E. R. Deacon, C.B.E., F.R.S., says: ‘We have learnt from the Journal that one of the most frequent sources of error is a poor horizon, and recent (my italics) experiments by our American colleagues have shown that much can be gained by fitting a more powerful telescope in the sextant so that stars can be used while the horizon is better illuminated. They showed that it is better to look at the star directly and to make the horizon the reflected image.’

Dr. Deacon has been misled. In the ‘old’ Discovery from 1925 onwards (and before that in various surveying ships, whilst on passage), in which I was Navigating Officer, the horizon was always brought up to the star. A ‘sounding’ sextant also was always used. This is lighter than an ordinary observing sextant and mine had a large, aluminium telescope, of much greater power than the star telescope of an ordinary sextant, and with a very large object glass.

It was, however, graduated only to minutes of arc. To overcome this the altitude was pre-set for every contact to fixed intervals of, I think, 5′; this had the additional advantage that if the time of the middle contact agreed with the mean of the others it formed an excellent check on the accuracy of the sights and could be seen at a glance.