to me, some rough drawings of which I lay herewith before the Royal Society. As the phenomena are only momentary, I cannot answer for the final accuracy of the drawings, nor have I been able to represent the softness of the gradations of shade.

2. In the drawings, the red end of the spectrum is to the left; the D line common to them all is the image of a slit about half an inch long, on which slit the light falls from an electric lamp, through the tube and chamber in which the vapours are produced. The lower part of the drawings would generally represent, therefore, the spectrum of the less dense vapours were the vapours at rest.

3. One of the phenomena referred to consists of what may be described as a unilateral widening of the line D: the side absorption, however, is much less dense than that of the line; it is bounded by D on one side and by a curved line on the other. Figs. 1, 2, and 3 will give an idea of this appearance in three stages as it is frequently actually seen, i. e. as the absorption travels up or down the line it widens as shown.

Fig. 1.

Fig. 2.

Fig. 3.

4. Figs. 4 and 5 give two variations sometimes observed-fig. 4 showing the darkening in the absorption and an increased steepness in the curve; fig. 5 the simultaneous existence of apparently different absorptions, all bounded by D on one side, but by different curves on the other, and being of different intensities.

Fig. 4.

Fig. 5.

5. Although, in the preceding drawings, I have represented this unilateral widening exclusively on the more refrangible side of D, I have observed it on the other, though scarcely so frequently.

6. Accompanying these appearances, but generally best visible when the absorption with curved boundary is visible on both sides of D, is a brilliant boundary replacing the mere change of shade.

7. At times the brilliant boundary is continuous across D, as shown in fig. 6; but I append figs. 7 and 8 to show that the phenomena on either side of D are independent of each other.

to me, some rough drawings of which I lay herewith before the Royal Society. As the phenomena are only momentary, I cannot answer for the final accuracy of the drawings, nor have I been able to represent the softness of the gradations of shade.

2. In the drawings, the red end of the spectrum is to the left the D line common to them all is the image of a slit about half au inch long, on which slit the light falls from an electric lamp, through the tube and chamber in which the vapours are produced. The lower part of the drawings would generally represent, therefore the spectrum of the less dense vapours were the vapours at rest.

3. One of the phenomena referred to consists of what may b described as a unilateral widening of the line D: the side absorp tion, however, is much less dense than that of the line; it is bounded by D on one side and by a curved line on the other. Figs. 1, and 3 will give an idea of this appearance in three stages as it frequently actually seen, i. e. as the absorption travels up or dow the line it widens as shown.

Fig. 1.

Fig. 2.

Fig. 3.

4. Figs. 4 and 5 give two variations sometimes observe showing the darkening in the absorption and an increased in the curve; fig. 5 the simultaneous existence of appare ferent absorptions, all bounded by D on one side, but by curves on the other, and being of different intensities.