

184 



GEOLOGY. 



[Sect. VI. 



ample opportunities of examining volcanic islands, and 

 perhaps volcanoes in eruption. With respect to the lat- 

 ter, he ought to record all that he sees : should the exact 

 position of the orifice be known, he might, perhaps, 

 observing some point in a cloud, measure with a sextant 

 to what height the fragments were shot forth, and the 

 height of the often flat- topped column of ashes. Having 

 surveying instruments, he ought to map, as carefully 

 as time will permit, any crater remarkable for its size 



Elie de Beaumont has 



depth, 01' peculiar form. M 



found that, owing to the fluidity of lava, streams never 



consolidate into a thick, moderately-compact mass, ex- 



cept on a surprisingly gentle inclination. On 

 above 2 or 3"", the stream consists of extremely irregular 

 masses, often forming a hollow vault within. Fresh ob- 

 servations on this point are much wanted in regard to 

 lavas of different composition. The measurements can 

 easily be made by a sextant and artificial horizon.* In 



M. Elie de Beaumout gives the foUowiiig directions (Memoires pour 

 servir, &c,, torn. iv. p. 173) : 



" The method I arn in the habit of employing for these kinds of mea- 

 surements is simple and easy, and a description of it may save useless 

 trouble to others. I place on the edge of the sextant, and behind the 

 fixed mirror, a small piece of white paper, in which there is a narrow 

 opening (ouverture etroite) corresponding to the axis of the telescope. 

 On the exterior surface of the paper a black line is drawn, perpersdicular 

 to the i>lan of the graduated circle, and passing through the centre of 

 the opening above mentioned. A quantity of mercury is poured into a 

 vessel sufficient to Ibrm a plane horizontal surface of a certain extent. 

 The telescope of the sextant is then directed vertically over the mercury, 

 and the image of the black line sought for. When this is found, I am 

 certain that the visual raj/ from the image In the mercury can only 

 deviate from the perpendicular^ in so far as the line is not without 



i, and the opening has a sensible size. These two sources of error 

 can be diminished so that the maximum of error shall not exceed a 



br 



T .-■ 



niinute. Being once certain of the verticality of the visual ray f. 'n the 



I 



I 



i 



gect 



tn^' 



i--"^ 



lO 



vU 



of 



En' 





str 

 des 



it 



s 



] 





siet 

 Han 

 evei 



colli 

 moi 



dec 



i!i 



m t 



SC 



tion 



m 



.Q*- 



oftl 



iv^. 



^ 



^ \~ 



^ \ 



toli 



f 



