excess of the alkali has acted in the way suggested by Dr. Zschimmer, and produced the assemblages of crystals described above.

It may be here remarked that these assemblages of crystals are very similar in growth to those formed by the crystallisation of potassium permanganate, as described and illustrated by Professor Lehmann on page 124 of the volume on Flüssige Kristalle to which reference has previously been made.

In order to try and reproduce the forms of crystallisation deposited on the lenses, a series of experiments was made with partial success.

As potassium (17%) was the chief ingredient next to silicon (69.8%) in the composition of the glass, solutions of pure potash with distilled water were made and poured on to the cleaned surfaces of several cover-glass lantern plates, and then drained off. The moistened sides were then covered with another cover-glass, but separated by means of small wooden diaphragms in order to allow a slight circulation of air. The glasses were kept in an upright position in a cupboard, which was frequently opened.

In a similar way, several other glasses with solutions of potash and caustic soda, mixed in the proportion of 3 to 1, were also prepared and treated in the same way.

After 10 or 15 days crystallisation in most cases began to set in, and the plates were examined and occasionally photographed. It was most interesting to observe, not only the different forms of crystallisation on the various plates, but also the different forms on the same plate.

In the few attempts made, the form of crystallisation shown in Plate 2, fig. 1, was not exactly reproduced, although in some instances a close approximation was reached.

In conclusion, it may be remarked that the deposit of crystals on the several lenses to which reference has here been made has in no instance been removed at the observatory. The objectives were always sent away to be cleaned, and the method of getting rid of the deposit was to rub the surfaces with material soaked in dilute sulphuric acid. It is, however, satisfactory to know that Messrs. Schott and Genossen no longer manufacture the particular kind of glass which has caused this trouble, and have erased it from their catalogues. The more recently made photo-visual lenses should not, therefore, in future be hampered with this defect.

Before submitting this paper to the Society, I sent it to Mr. H. Dennis Taylor, whose investigations on the theory and practice of photo-visual lenses are so well known. He has kindly written an "Addendum," which is here attached, and which will be found a valuable contribution to the subject of the present paper.

Addendum by H. Dennis Taylor.

Notes on Dr. Wm. Lockyer's paper on the Permanency of some Photo-visual Lenses.

Dr. Lockyer kindly submitted to me his paper before publication, with a view to my adding any remarks of my own on this interesting question.

In the first place, I would like to point out that the dates of mounting the instrument given by Dr. Lockyer are not the dates on which the objectives were put together, so that, correctly, at least 12 months should be added to the periods given, on the average.

Dr. Lockyer says that he had not been able to notice any deterioration of the optical performance of any of these objectives after being cleaned up, but that no special tests were employed.

But after cleaning at Messrs. T. Cooke & Sons' works, the objectives were carefully examined and tested for optical deterioration and none whatever was noticed. One would have thought that the crystallising out of some of the alkaline constituents of the glass would have left a network of channels in the glass surfaces, and for the presence of such small inequalities, tests by reflected light should be more appropriate, but we could not see anything. Of course, it is quite possible that the repetition over many years of such crystallisation, supposing it does repeat itself, might deteriorate the surfaces of the lenses sufficiently to cause optical mischief, but so far it has not been apparent.

With regard to the small bundles of glistening crystals forming upon the inner surface of the front lens of baryta light flint, and shown in fig. 2, Plate 2, and figs. 1, 2, and 3, Plate 3, I would like to observe that these never seem to be at all fast to the surfaces, and can always either be blown off or brushed off with a delicate camel-hair brush; and after so brushing them off, the surfaces are just as clean as before, excepting for a little dust that may have found its way in.

As regards the much more noticeable crystallisation on the inner surface of the back lens, I have never had any difficulty in rubbing it off with a soft rag dipped in very weak sulphuric acid, which, after thorough washing off, leaves the surface just like new. The amount of this crystallisation is largely dependent on the dampness or dryness of the situation in which the objective is used, and it would most probably never form at all could we keep the objective always in a dry atmosphere. Moreover, it depends also upon the particular type of crown glass used. I must here point out that the crown glass with a larger amount of alkali than usual, referred to by Dr. Lockyer as O 374, was employed in the 12-in. aperture, 3 aperture, and one of the two 3-in. aperture O. G.'s, while the other two, of 4-in. aperture, were made out of a borosilicate crown glass, type O 599, of Messrs. Schott's catalogue,