Very different was the case with a full moon, especially in the long, often cloudy, nights of winter. Then a vigilant watch had to be kept from sunset to sunrise, so that in case of a short break in the clouds the moon might yet be observed. Such a watch was the severest (with one exception) that an assistant had to undergo. His fourth day would then resemble his second, and with the fifth day a second cycle of his quartan fever would commence, the symptoms following each other in the same sequence as before. Such a routine carried on with iron inflexibility was exceedingly trying, as it was absolutely impossible for an observer to keep any regularity in his hours of rest or times for meals. This routine has been considerably modified by the present Astronomer Royal, partly because the instruments now in regular daily use are five instead of two, and partly because a less stringent system has proved not merely far less wearing to the observers, but also much more prolific of results. It was impossible for a man to be at his best for long under the old _régime_, and from forty-six to forty-seven has been an ordinary age for an assistant to break down under the strain. One point in which the observing work has been lightened has been in the discontinuance of the altazimuth observations at the full of the moon, another in the shortening of the hours of the transit circle watch; and a further and most important one in the arrangement that the observers with the larger instruments should have help at their work. The net result of these changes has been a most striking increase in the amount of work achieved. Thus, whilst in the year ending May 20, 1875, 3780 transits were taken with the transit circle, and 3636 determinations of north polar distance; in that ending May 10, 1895, the numbers had risen to 11,240 and 11,006 respectively, the telescope remaining precisely the same. One principle of Airy's rule still remains. So far as possible no observer is on duty for two consecutive days, but a long day of desk work and observing is followed by a short day of desk work without observing.