Dr Daubeny on the Spring at Torre del Annunziata. 223 



at other times heaped together confusedly without any cement- 

 ing material. 



The beds themselves are of vai'ious shades of colour, grey, 

 red, and black, according to the predominant tint in their com- 

 ponent parts. 



In the lower portion of the cliff these beds are for the most 

 part tolerably distinct ; but in the upper they are so blended to- 

 gether, as scarcely to be distinguishable one from the other. In 

 the midst of these strata of rapilli occurs a bed of grey tuff, 

 more compact than that covering Herculaneum, and in parts 

 nearly as much so as the generality of that of Posilippo. It is 

 of irregular thickness, swelling out in one place, and contract- 

 ing in another. Its greatest thickness is about five feet. It is 

 seen to terminate abruptly on both sides ; on the east a little be- 

 yond the New Pump-room ; on the west immediately behind that 

 building. In the latter place it is interrupted apparently by a 

 heap of large blocks of volcanic matter, which are seen abutting 

 against its side. About ten feet farther to the west, however, 

 and precisely on the same level, we observe the commencement 

 of a narrow thread of the same species of tuff as before, which 

 pursues a similarly horizontal course along the cliff to the west, 

 gradually thickening until it acquires nearly an equal bulk to 

 that of the bed of tuff first noticed. Hence it would seem like 

 a continuation of the former bed, which had been broken off by 

 the heaps of detached fragments of lava before noticed. 



This second bed of tuff, however, after pursuing a tolerably 

 regular course for about thirty feet along the cliff, breaks off as 

 suddenly, and as abruptly as the former, so that both appear to 

 be in a manner inclosed within, and surrounded by, the strata 

 of rapilli. At a still lower level are one or two beds of tuff si- 

 milarly constituted, though less distinct than the two already de- 

 .scribcd, and these, in like manner, after continuing in an hori- 

 zontal direction for a few feet, either break off abruptly, or are 

 gradually lost by intermixture with the surrounding beds of ra- 

 pilli. That these tufaccous strata owe their superior compact- 

 ness only to the weight of the beds of loose rapilli superincum- 

 bent, and have been formed on dry land by the mere action of 

 rains or torrents upon volcanic materials, appears from the occur- 

 rence, in a stratum of rapilli and volcanic sand eight or nine feet 



a2 



