^{page 53 note *} The Royal Society has not ceased its work, and holds itself pledged to finish what it has undertaken, namely, to catalogue the scientific papers of the nineteenth century.

^{page 55 note *} A third venture might here have been referred to, namely, the Répertoire hibliographique des sciences mathématiques, a card catalogue following the system of classification adopted by the Revue. Between 1894 and 1905 fifteen hundred cards have been sent out, and the present rate of issue is only one hundred cards per year.

^{page 58 note *} The list lays itself open to criticism, but let him throw the first stone who knows the difficulties of organising the machinery necessary to produce not one year-book but the whole seventeen annual volumes of the International Scientific Catalogue. For one thing, the director of such an undertaking will fi d his numerous regional-bureaux at sixes and sevens as to where the line of exclusion is to be drawn, and consequently elementary journals will be inserted in one land and shut out in another. Then, of course, journals have an appreciable death-rate, and no list can be complete and accurate for long: that here used is the first drawn up. In view of what lias been already accomplished, alike by the Council and by the Director, the manifest duty of all scientific men is to lend what aid they can towards now perfecting the Catalogue in detail.

The number preceding the title of any serial is the consecutive number which distinguishes it in the above-mentioned List of Journals.

^{page 59 note *} Of those who have assisted me in obtaining details, I desire specially to thank Professor Gibson of Glasgow, and Mr Hardy, the worthy librarian of this Society.

The London libraries which are of importance as regards mathematics are six in number, namely, those of the British Museum, Royal Society, Mathematical Society, University College, South Kensington Science, Patent Office. Some of these possess more serials than any one of the Scottish libraries ; but the same general state of aifairs prevails. Overlappings repeatedly occur, serials possessed are frequently imperfect, and the full collection of the testlist cannot be furnished forth by all the libraries put together.

The library which possesses representatives of the greatest number of serials is that of the Mathematical Society: unfortunately not more than half a dozen of the sets are complete, and the books are miserably housed. To provide proper accommodation, to complete the sets already represented, and to supply the missing sets, would be a splendid ambition to set before the little band of members resident in London. The serials contained in the British Museum, though somewhat fewer in number, are wonderfully complete and excellently cared for.

Meanwhile it is desirable that a leaflet should be prepared containing a full list of serials, and showing which are obtainable in London, and where. For those which are more elementary, the library of the Mathematical Association might be utilised.