Whereas it is doubtless true that the two largest contributions to the errors of sights at sea are those referred to by Mr. Sadler in his Prefatory Note to Captain H. H. Shufeldt's paper on Precision Celestial Navigation Experiments, it cannot, I think, be gainsaid that the principal avoidable error in sights is due to faulty use of the sextant, especially in respect of collimation error.