VOYAGE TO ALGERIA 175 



or limestone water to the open air partially softens it. A 

 specimen of the Redboume water exposed bj Professors 

 Graham, Miller, and Hofmann, in a shallow basin, fell 

 from eighteen degrees to nine degrees of hardness. The 

 softening process of Clark is virtually a hastening of the 

 natural process. Here, however, instead of being per- 

 mitted to evaporate, half the carbonic acid is appropriated 

 by lime, the half thus taken up, as well as the remaining 

 half, being precipitated. The solid precipitate is per- 

 mitted to sink, and the clear supernatant liquid is limpid 

 soft water. 



We returned to the real mouth of St. MichaeFs Cave, 

 which is entered by a wicket. The floor was somewhat 

 muddy, and the roof and walls were wet. We soon found 

 ourselves in the midst of a natural temple, where tall col- 

 umns sprang complete from floor to roof, while incipient 

 columns were growing to meet each other, upward and 

 downward. The water which trickles from the stalactite, 

 after having in part yielded up its carbonate of lime, falls 

 upon the floor vertically underneath, and there builds the 

 stalagmite. Consequently, the pillars grow from above 

 and below simultaneously, along the same vertical. It 

 is easy to distinguish the stalagmitic from the stalactitic 

 portion of the pillars. The former is always divided into 

 short segments by protuberant rings, as if deposited peri- 

 odically, while the latter presents a uniform surface. In 

 some cases the points of inverted cones of stalactite rested 

 on the centres of pillars of stalagmite. The process of 

 solidification and the consequent architecture were alike 

 beautiful. 



We followed our guide through various branches and 

 arms of the cave, climbed and descended steps, halted at 



