Prof. How on the Mineralogy of Nova Scotia. 133 



his mineral to lose water with great facility, and that the spe- 

 cimen I analyzed was partly opaque and had been exposed for 

 an unknown length of time, it may seem that I should be led 

 to the view that centrallassite is really gyrolite. This is sup- 

 ported by the fact of its structure being lamellar. Anderson 

 says of gyrolite, " its crystallization in plates and not in needles, 

 its high lustre, and the ease with which it loses water distin- 

 guish it completely from dysclasite " (okenite). To this it may 

 be added that while this last mineral gelatinizes readily with 

 acid, centrallassite does not ; and though Anderson, curiously 

 enough, does not allude to this important character, and the 

 same is true of myself on the occasion above cited, I now find 

 that the mineral I then described as gyrolite does not gelati- 

 nize, but resembles centrallassite in giving slimy silica. Not- 

 withstanding all this, I am still held to claim individuality for 

 centrallassite, because I proved in my paper on the three mi- 

 nerals, at least to my own satisfaction, that my mineral does 

 not lose water readily. I showed that on taking portions from 

 the interior of the mass, some " without lustre, chalk- white, 

 and in which a radiated structure was sometimes, but not inva- 

 riably, quite obvious," in two cases the percentage of water 

 was found to be 12*29 and 12*25, which is a little above the 

 average, 11*41, and was supposed to be accounted for by ab- 

 sorption of water by the white powdery form, because, as I 

 find from my notes, a direct experiment was made with regard 

 to the loss and absorption of water. Some fragments, selected 

 as being nearly pure centrallassite, were powdered and weighed 

 at once, being placed over H 2 S0 4 ; the loss in forty hours was 

 at an end and amounted to 0*53 per cent. ; and on exposure to 

 the air for two days the gain by the dried powder was com- 

 plete and equalled 0*73 per cent ; on redrying over H 2 S0 4 , 

 the loss on the original weight was the same as before, and on 

 reexposure for two hours only or less the gain was 0*59 per 

 cent. 



My opinion, from the whole evidence, is that, from the cor- 

 respondence of okenite, gyrolite, and centrallassite in consti- 

 tuents, and their approximate quantitative composition, together 

 with the occurrence of a few prismatic needles in tufts along 

 with the pearly plates, or in close proximity, noticed by myself 

 formerly in the case of the latter mineral (" two very small 

 tufts of divergent silky transparent needles seen in the interior 

 of the nodule, having the blowpipe characters of the lamina?," 

 loc. cit.), they are closely related indeed, but the merging of 

 one in either of the others as a species is not admissible. 

 Taking the formula I now substitute for my old one for cen- 

 trallassite, that of Anderson, which I have shown anew to be 



