The material which I wish to present today is set out in Royal Observatory Bulletin Number 66, of which I have a few advance copies with me. The paper describes measures of about 700 stars in an area about 1° square. These are nearly all the stars down to a limit of V =14m5 in the area. Proper motions were measured by comparison between two plates taken on 1912 Jan. 21 and 1960 Nov. 16–17. The earlier plate had an exposure of 5 hr, which is exceptionally long for astrographic plates of those times. The measures of Ax and Ay (the shifts in units of one-tenth of a micron in the directions of right ascension and declination) were reduced by constructing plate constants such as would make the least-squares motion of the stars with B – V < 0 zero. In the reduction 104 such stars were used. The standard errors of measurement are σ(x) = ±15·9 and σ(y) = ±14·9, assuming that these stars really are members of the LMC and exhibit no relative proper motion. By using these plate constants, values of Ax and Ay were computed for all the remaining stars. The stars with 0·5 < B – V < 1·5 (apart from cepheids and a few reddened stars) are very plainly shown by their proper motions to be nearly all field stars, and indeed their motions can be satisfactorily explained by supposing them to be galactic stars showing the parallactic motion, galactic rotation, and a velocity ellipse appropriate to galactic stars. Stars with B – V > 1 · 5 cannot be determined to be members of the LMC on proper motion alone, but other considerations suggest that this is so. For example, their x, y distribution in the field follows the same pattern as that shown by the very blue stars, whereas the field stars (0 · 5 < B – V < 1 · 5) are uniformly distributed over the plate.