With the purpose to determine the relationships between noise and various hearing losses, the author studied the masking phenomena of the pure tone by air conduction and bone conduction in the presence of noises composed of various frequency components (although these were not completely white noise); and obtained the following results. 1. The masking curve of the normal ear in the presence of a given amount of noise approximately parallels with the spectrum curve obtained by the analysis of the noise employed. 2. The relationship between the change in the level of noise and the consequently appeared transition of the threshold under masking in the normal ear, differs according to differences in the frequency in the starting point of masking, but it changes in a straight line once the masking starts to take effect. 3. By drawing a masking curve under a given amount of noise of 108 ears with hearing loss, this curve was compared with the standard masking curve, the curve drawn from the average value of hearing threshold of 20 normal ears in the presence of 60-phone noise. As the results it has been found that: A) Even if the difference between the two curves is under 5 db or as much as 10 db, as long as there is one or two frequencies only; namely, those that show the curve (the curve of unisolated type) that is about parallel with the standard curve occupy as much as the whole of those with conductive deafness, about 76.4 per cent of those with perceptive deafness and 70 per cent of those with mixed deafness. B) Those that show a relatively big difference between the two curves, for example, a curve (curve of isolated type) that shows a difference as much as over 10 db, can be found in perceptive and mixed deafness. Four cases assumed to have retrolabyrinthine deafness show an uuexpectedly high value of the threshold along with the increase in the noise level, and also the changes in the their threshold transition are not in a straight line. 4. In order to study the relationship between the change in the noise level and the change in the threshold transition, the masking curve (standard curve) that shows the transfer of the normal ear threshold due to noise was compared with the masking curve of deaf ear. A) The curve (type I) where the transition of the threshold of deafness does not move so much as it will intersect the standard curve can be seen in 79 per cent; and next, the curve (type II) where the transition of the threshold moves as much as it intersects the standard curve can be observed in 21 per cent, showing the type I curve in a decidedly greater percentage. B) In observing the rate of appearance of the type I curve according to the sorts of deafness, namely, conduction deafness, perceptive and mixed ones, it is greatest in conduction deafness, followed by mixed one, and perceptive one in the order mentioned. 5. The curve (type A) which moves in a straight line about parallel with the standard curve from the time when the threshold of deaf ear starts to move under noise can be seen in 83 per cent; and the curve (type B) which gradually approaches the standard curve after the initiation of the threshold movement can be observed in 5 per cent; and the curve (type C) that moves rapidly away from the standard curve after the initiation of the threshold transfer can be seen in 12 per cent, indicating a preponderance of type A curve. 6. The masking curve of bone conduction in various ears, the same as in the case of air conduction, moves approximately in parallel with the curve obtained by the analysis of noise effect, and by the chages in the noise level the threshold moves in a straight line. Generally the rise in the threshold value is less in bone conduction than in air conduction, and the masking curve in conduction deafness somewhat coincides with that in normal ears. Moreover, the rate of appearance of type II curve is great in perceptive deafness.