recent years at least, falling below one fifth. From one third to one fourth may be taken as a general average. The causes for these fluctuations are not difficult to trace, although at first sight they are not very evident. The migratory character of the association is to a considerable extent the main reason. It has been the custom for many years to meet somewhat alternately in Eastern and Western cities; and the fact has been conspicuous that the Western meetings have usually been "off years" in attendance. At the same time, if the membership from accessions in the previous year, at a large meeting on the seaboard, has been high, the smaller attendance makes a marked reduction in proportion to the membership. Other circumstances have at times caused increase or diminution in attendance in special ways. Thus the Springfield (Massachusetts) meeting of 1895 v/as held so late in August that many teachers could not attend it and be at home in time for the opening of their work at the beginning of September, and this palpable error was probably responsible for an unusually small attendance at a pleasant and convenient place. The Detroit meeting of 1897 was overshadowed by the meeting of the British Association at Toronto in the following week, as some who could not attend both occasions chose the latter as the more remarkable. On the other hand, the great Philadelphia meeting of 1884 owed its unequaled attendance, both actual and proportional (twelve hundred and sixty-one out of a total of nineteen hundred and eighty-one), to the presence of a large number of British members who came from the meeting of their association, held the week previous in Montreal, just as some four hundred American members went last year to Toronto.
Whenever the place of meeting is a large city, and particularly if it be one containing important scientific institutions, like Washington, Philadelphia, or Boston, both the membership and the attendance are greatly increased by the addition of many "local" members. Some of these remain permanently, while others drop off in the course of a few years; but the result is a net increase, though subject to many fluctuations, as we have already noted.
One fact is apparent, and very gratifying, from a study of these figures in detail—to wit, that notwithstanding a certain impression to the contrary in some quarters, the association has, since the civil war, maintained a steady though varying growth in numbers, and (when several years are taken together) a fairly uniform proportion of attendance to membership. At the recent meeting, the register of those present showed a little over nine hundred, while the total membership must be considerably above two thousand, and doubtless larger than ever before.
There seems no foundation, therefore, for the idea expressed
