Against the vaidyans’ figures we may place the statistics of the Madras Hospital, even so long ago as 1903, and before a rigid system of antisepsis had been introduced. Recoveries numbered 96 per cent., poor results 2 per cent., and failures 2 per cent. These figures would indicate that the coucher was losing 60 per cent. more eyes than the hospital did even then. If the vast number of eyes submitted to couching be taken into account, this 60 per cent. of avoidable loss totals up to a staggering figure. Nor must we disregard the fact that, even amongst the successes, the average vision obtained is greatly in favour of the Western surgeon. The table showing the causes of failure will repay a careful study. The figure for iritis and iridocyclitis comes to 35·76 per cent. of the _total_ number of cases; glaucoma accounts for 11·05 per cent., imperfect dislocation of the lens for 8·94 per cent., retinal detachment for 3·53 per cent., optic atrophy (including one case of optic neuritis supervening as a septic complication of the operation) for 2·59 per cent., retinitis pigmentosa and retinitis punctata albescens for 0·49 per cent., retinochoroiditis for 1·41 per cent., vitreous opacities (admittedly a very vague term) for 1·18 per cent., and failure due to operation on a congenitally imperfect eye for 0·23 per cent.; 3·53 per cent. are, unfortunately, unaccounted for owing to deficiencies in the notes. In the great majority of cases ruined by iridocyclitis the inflammation made its appearance within a few days after operation; but there were instances in which this complication was delayed for a long period. Our notes show three cases in which it came on from one and a half to three years after operation, one case after seven years, and one after ten years. There are also a few doubtful cases in which a history of three or four months of useful vision preceded the inflammatory attack. In one case at least, sympathetic ophthalmia would appear to have destroyed the other eye two years after operation. Similarly, it was found in most cases of glaucoma that the access of high tension came on within a few days of operation. There were six exceptions to this rule, three commencing from two to ten months after the couching, one five years, one six years, and one fourteen years after. From a clinical point of view, the cause of the onset of glaucoma in these cases is obscure. Many of them appear to be associated with iridocyclitis, but we must leave this matter for the present. We shall have occasion to deal with it much more fully under the heading of pathology.