6 EXCURSIONS IN MADEIRA 



of those horizontal fissures created by the moisture eagerly 

 attracted from a marine atmosphere — an apparent stratification, 

 which is merely superficial, and only wrought by a long contact 

 with the air. The opposite drawing (Plate 2), which was carefully 

 made on the spot, will convey a better idea of the appearance 

 of these rocks, which evidently supplied the stone for the 

 aqueduct that strides over them^ 



During this short stage of the ascent, the anthemis hispanica 

 presented itself, but was confined to a very small space; the 

 inula odorata extended a little further; the convolvulus arvensis 

 confined itself to the bottom, and although I did not discover 

 a trace of the genus during the remainder of the ascent, yet on 

 arriving at the highest point, in the most exposed situation possible, 

 I found the convolvulus tricolor. The scabiosa srtccisa was thinly 

 scattered at the foot of the first hill ; the sisymbrium palustre also 

 confined itself to the bottom of the hill ; the inentha arvensis grew 

 at the foot, and frequently presented itself until half way up 

 the ascent, where it totally disappeared. About midway, there 

 were some dwarfish tufts of the ulex europaus; and a great 

 profusion of the genista viscosa, the euphorbia dendroides, and 

 the atractylis humilis — the former plentifully, the latter thinly 

 scattered, were found, with the carduus eriophorus and the 



'' L'Eveque writes, that the great arch of the aqueduct is 100 feet, three inches, 

 wide between the pillars, and that, from the keystone of the arch to the bottom of the 

 rivulet, its height is nearly 206 feet, and 214 to the parapet. The plan published by 

 Wells in 1792, (from that presented to the Marquis of Pombal,) makes the height of 

 the grand arch 226| English feet, and the width 108. I made the height, from the 

 bottom of the rivulet to the parapet, 222 English feet, by a barometer of Fortin's, 

 which marked on the parapet wall 750.66, (thermometer 20f, thermometer detached, 

 20^, centig.) and on the side terrace under the great arch, (5 feet above the bed 

 of the rivulet,) 756.50, (thermometer 20|, thermometer detached 20f.) The length 

 of the work at the valley of Alcantara is 2873g feet; the whole length of the 

 aqueduct (from its source at Canessas to Lisbon) is 56.380 feet, following sixty-five 

 windings and dingles. 



