to measurement. A great desideratum in this branch is a new general catalogue of double stars; for such a work would greatly facilitate reference, and save the trouble of searching through different lists in order to identify an object. Burnham has given some practical hints on double-star work in the ‘Sidereal Messenger,’ and his remarks are reproduced in that excellent work ‘Astronomy for Amateurs.’ As to variable stars, some of these permit of naked-eye estimation, others need a field-glass, and there are some which require to be followed in a good telescope. The observer who enters this department may either desire to find new objects or to obtain further data with regard to old ones. If the former, he cannot do better than watch some of the suspected variables in Gore’s Catalogue of 736 objects, published by the Royal Irish Academy. Whether suspected or known variables are put under surveillance, the plan of comparison will be the same. Several stars near the variable in position, and nearly equal in light, should be compared with it, and the differences in lustre, in tenths of a magnitude, recorded as frequently as possible. The extent and period of the variation will become manifest by a discussion of the results. The comparison-stars should of course be constant in light, and, if naked-eye stars, they may be selected from the _Uranometria Nova Oxoniensis_ or ‘Harvard Photometry.’ If telescopic stars are required, then recourse must be had to comprehensive charts such as Argelander’s _Durchmusterung_, which includes stars up to 9½ mag. Variable stars of the Algol type are especially likely to escape recognition, as they retain a normal brilliancy except during the few hours near the time of a minimum.