ABICH ON THE NATRON LAKES OF THE ARAXES. 29 



and also on the under surface of the crust where it adheres to the 

 clay bottom. 



The salt mentioned above as floating on the surface of the water is 

 only distinguished from that just described by less compactness in the 

 union of the layers which are scarcely a line in thickness, and being 

 separated from each other by small horizontal spaces, have a more 

 laminar aspect. It is also nearly pure white, with a pale rose-red 

 tinge. 



The salt from the bottom of the lake had the following composi- 

 tion : — 



Sulphate of soda (Na S) / 4*44 



Carbonate of soda (Na C) 18-42 



Chloride of sodium (Na Cl^) 1-92 



Water M8 



Manganese and magnesia traces 



98-96 



This salt ignited in the platina capsule retained its crystalline cha- 

 racter and pearly lustre to the point of fusion, but assumed the colour 

 of chloride of silver. It requires a continuous and strong ignition 

 over the spirit-lamp to cause it to fuse. 



The salt from the surface of the lake had this composition : — 



Sulphate of soda 80-50 



Carhonate of soda 1 0*09 



Chloride of sodium 1-62 



Water * 5 5 



Mao-nesia and Mansranese traces 



98-82 



The almost entire absence of water of crystallization, as well as the 

 physical peculiarities of these salts, are remarkable facts, and show 

 that the sulphate and carbonate of soda in the anhydrous condition 

 can, in favourable circumstances, combine in a double salt. The 

 combination is not subject to deliquescence, and it would appear has 

 no tendency to attract moisture, for masses of it placed for a long 

 time in damp situations yield no more water than those previously 

 exposed to a strong heat. The affinity of the two salts, conducing 

 tothe formation of an anhydrous double-combination, must be stronger 

 in the given conditions than that of water to either ; for in no place 

 could any other crystallizations be discovered in the fluid. Hence 

 this salt forms in reality a new mineral species, most nearly allied to 

 the Thenardite, and which might not inappropriately be named Ma- 

 kite, as the lake is situated in the territory of the Chan of Maku. 



The water of the lake itself has entirely the colour of a concen- 

 trated solution of sulphate of manganese-protoxide*. In 100 parts 



* The peculiar colouring principle is not known to me at present. I have ex- 

 amined it for Bromine, bur without finding any trace of this substance. I would 

 conjecture that it is some vegetable matter. 



VOL. v. PART II. E 



