required subsequently; afterwards the diminution is slower, or in some cases even an increased quantity of oxygen may be required, from changes taking place in the water causing the production of a larger quantity of readily oxidisable matter. This therefore gives no indication of the weight or actual quantity of organic matter present The weight of organic matter ascertained by experiment however, indicates in some cases a rather rapid diminution at first. But this is only to a small amount, except in the case of highly decomposable or putrefying liquids, such as sewage or mixtures containing much sewage. Calculation will show that the loss of weight of organic matter in mixtures Nos. 1 and 2 is less than would have been sustained by the constituent proportion of sewage water in them. In No. 3 probably the vegetable matter of the Tank water added caused the more rapid and extensive decomposition.

The loss of weight in the mixture No. 3 is 3.4 grains in 8 days, being fully more than half of the original amount; in Nos. 1 and 2 it is only only 1.8 grains and 0.3 grain respectively.

But the river water at no time could contain anything like the proportion of sewage that these mixtures did, such as one-fifth, oneeighth or even one-twelfth of sewage, the smell alone of such mixtures, makes the supposition quite inadmissable; besides a comparison of the size of the river with the amount of drainage of the town would show that such a proportion was quite impossible. The amount of liquid discharged by the drains compared with the volume of water in so large a river must be insignificant.

But instead of citing results of my own, which if incorrect may be supposed to be all equally incorrect, it may carry more weight to quote the results of others. The older determinations of organic matter are generally of no value whatever, and I shall refer only to the most recent and trustworthy. I have already quoted Dr. Frankland's results with the London waters, but as all these are of water filtered for distribution they may be considered not quite comparable. Another example I shall adduce from the paper of Lawes and Gilbert in the Journal of the Chemical Society already quoted. They give tables of the composition of the Rugby sewage from May 1862 to October 1863, shewing that it contains in solution from 7.6 to 8.35 grains per gallon, and also a statement of the amount in the River

Wandle before receiving the Croydon sewage which is 1.44 grains per gallon, and after receiving it which is 2.08 grains per gallon. According to this the estimates of organic matter to the extent of 8 or 10 grains per gallon in the Hooghly water during May and June shew that it contains fully more than the liquid part of the Rugby sewage, and this in a tropical country.

I do not wish it to be understood that I maintain the perfect accuracy of my own results. The oversight in not examining the samples speedily enough after collection must be admitted, though from all that I have been able to learn from the experiments instituted for the purpose, the error cannot be a great one. There was no great delay in examining the water of the hot season,-about ten or twelve days, and this caused by the time and attention taken up in examining the influence of the tides in numerous samples. There was greater delay with the water of the rainy season, probably about a month with the first samples in July, about a week or 10 days with those of August. This was caused by waiting for the settlement of the very finely divided clay, the presence of which was very unfavourable to the accurate estimation of the organic matter. Recently I have found that the addition of a small quantity of hydrochloric acid causes the mud to settle so rapidly that the water may be filtered clear in course of a few hours: solution of potash or soda and milk of lime do the same, but the water cleared by these reagents seems to contain a different proportion of organic matter than that cleared by simple subsidence. It is of less importance, as the question at present is not respecting the water of the rainy season. The samples of December and February water circumstances prevented me from proceeding with, and they were preserved in stoppered bottles and probably not much done with them till April. The results are consequently more doubtful, though I do not suppose that they are very wide of the truth. As the season advances, should circumstances admit of it, I shall not fail to repeat the analyses, in order to get unobjectionable results.

The observations made during the last month enable me to add a little to my former statements respecting the effect of the change of seasons on the river water. The increase of organic matter from the rains seems to be chiefly of the more soluble and putrescible kinds; as

the season has advanced, the fetid smell has materially diminished. This is indeed to have been expected: the soil has been washed comparatively clean, and there is less of such matter to wash away.

The only possible way in which my results as to the small quantity of organic matter in the water of the hot season (supposing there is no great error in the analysis) can be reconciled with the results of those analyses that give it as equal to 8 or 10 grains per gallon, would be to suppose that the water at that season contains a large quantity of organic matter having no very offensive smell, but capable of very rapid decomposition, so that about th to 9th of it would be lost during the first two weeks. Without denying the possibility of this, I can only say that I know of nothing that makes it probable that such is the case, while I have already given reasons for believing that no such state of matters exists. Further observation and experiment can alone decide the question beyond doubt; while I may remark that if such be the case, it will be a fact well worth noticing and establishing.

It may also be observed, that as in the case of supplying towns the water must always be stored for a time in tanks or reservoirs, it is a point of some importance to note the changes which it undergoes by keeping in these circumstances. I have made some observations in the course of these enquiries suggestive of further investigations on this subject, and which may also have a bearing on the purification and preservation of such waters, a subject which has lately been occupying much attention in England. It is obviously a possible thing that one water may be putrefying but its putrescibility nearly exhausted, while another may be highly putrescible, and yet its actual putrefaction may be only about to commence. As regards the preservation of waters too, it is one thing to keep them in stoppered bottles, and another thing to keep them in tanks. It seems to me questionable if they improve in tanks as they do in glass bottles. It is by following out such inquiries that advance in knowledge of such subjects is attained, and in the present case the activity of chemical changes produced by the high temperature and the regularity of the seasons are in no small degree favourable for carrying them to a successful result.

Kashmir, the Western Himalaya and the Afghan Mountains, a geological paper by ALBERT M. VERCHE RE, Esq., M. D Bengal Medical Service, with a note on the fossils by M. EDOUARD DE VERNUEIL, Membre de l'Acadèmie des Sciences, Paris.

(Continued from page 203, of No. III. 1866.)

CHAPTER III.-Cursory Survey of the several chains of the Western Himalaya, the Afghan mountains and their dependencies. Preliminary geological mapping of the Western Himalayan and Afghan Ranges.

59. It is intended, in this chapter, to give, in as few words as possible, an idea of the general geology of the several portions of the Western Himalaya, the Afghan mountains and their respective dependencies. In doing so, I have availed myself of all sources of information which have been opened to me; I have, however, been sadly in want of the help of a more extended library, and I have never seen some excellent works which would have much improved this chapter, if they could have been consulted. I need therefore hardly say that it is a most superficial of surveys; but I hope nevertheless that it may be found to contain a few interesting observations and some new matter yet unpublished. Such as it is, it will enable us to sketch at least the first preliminaries of a geological mapping of the Himalayan and Afghan Ranges; and also to attempt, in the last chapter, to draw the history of the mightiest mountainous mass of our globe.

By reference to the map and and to the long Section (Sect. G) it becomes evident that the Himalayas are a succession of more or less regularly parallel chains, having a general N. W. to S. E. direction. Between the chains are situated valleys which are elevated above the sea in proportion as one nears the centre of the mountainous mass : thus the Rawul Pindie plateau, between the Salt Range and the SubHimalayan hills, is about 1700 feet high; Poonch valley, between the Sub-Himalaya and the Pir Punjal chain, is under 4000 feet; Kashmir between the Pir Punjal and the next chain (called in the map Ser and Merchain), is above 5000; Ladak between the Ser and Mer chain and the Kailas chain is 10,000 to 11,000; Nubra and the valley of the Shayok,

between the Kailas and Korakoram chains is a plateau nearly 15,000 feet high. It is probable that on the other side of the Korakoram chain the elevation diminishes and that the Aksai chain and the valley of the Yarkandkash river, between the Korakoram and Kuen-Luen chains, are about 10,000 feet high; beyond the Kuen-Luen is the province of Kotan which has been satisfactorily determined by its vegetation to be no more than 5000 feet high.

We have therefore a series of steps rising from the plains of the Punjab to the high plateau of Little Thibet, and descending from Little Thibet towards Turkish China. These steps are supported by parallel chains or walls which tower by some thousands of feet above the plateaux which they support. These chains offer a considerable impediment to the flow of rivers towards the plains, and most rivers have a considerable course parallel to the direction of the chains, before they can find a gap to pass through.

The Afghan mountains present the same arrangement as the Himalayas; the direction is from the N. E. to S. W. the direction of parallel chains is less well marked than in the Himalaya, but this is probably due to the little which is correctly known of the topography of these mountains. The plateaux are similarly graduating: Bunnoo being about 1200 feet above the see, Kabul 7000 feet, Kaffiristan higher, whilst the plateau of Koonduz, on the other side of the Hindoo Koosh, slopes gradually towards the west. This arrangement by plateaux is the same as is seen in the Andes with their high central chain and their plateau between that chain and the Cordilleras.

From the hypothesis, advanced in the next chapter, of the manner the Himalayan and Afghan mountains were upheaved, we will deduct which of the lower hills belong to the Afghan and which to the Himalayan mass, and I will therefore not discuss this subject here, as it would but lead to useless repetitions. I shall begin with the hills which one first meets crossing out of the alluvial plain of the Punjab, as he travels north from Mooltan; and I shall take the parallel regions of the Himalaya one after the other, noticing as I go on whatever little I know of the geology of the Afghan mountains in the same latitude.

60. In latitude 32° 10', longitude 70° 50' to 71° 20' rises the double chain of the Kafir Kote range or Rotta Roh and the Sheikh