On the Structure of the Delta of the Ganges. 329 



scarcely explicable on such a supposition. Thus the kankar 

 at Ghazipoor is described by the Rev. Mr. Everest as " re- 

 sembling much in the shapes it affects, the flints of the chalk 

 strata. The bank of the river (the Ganges) shews a section 

 of a layer about four feet thick, formed of these pieces. 

 This layer is gradually lost in the clay above and below. In 

 some places the layer or seam is double, with an intervening 

 bed of clay, through which the lower kankar bed is said 

 to branch off and join the higher. The beds are farther 

 similar to those of chalk flint in the manner in which the 

 loose pieces are imbedded in the clay, and in the layers being 

 composed of detached pieces." The origin of the kankar is 

 proved to be very modern, from the occurrence of existing 

 fresh water shells, small planorbes, &c. in the clay on which 

 it rests. The clay itself is calcareous. The kankar of Mir- 

 zapoor is described by the same observer as exhibiting like 

 phenomena. In the sections exposed by the bed of the 

 Jumna the kankar is found similarly in layers, between beds 

 of clay and calcareous marl, and in every instance I have 

 been able to find it, has been associated with mechanical de- 

 posits containing carbonate of lime. Now it appears some 

 presumption against the whole of these vast collections of 

 nodules being due to calcareous springs, that recent though 

 their formation has been proved to be, there are no inde- 

 pendent indications of the existence of the springs whence 

 they could have been derived. So universal is the preva- 

 lence of kankar, that it is scarcely possible to believe the 

 whole traces of its originating sources have been obliterated. 

 Farther, from the appearance of the beds with which the 

 kankar is associated, it is evident that no springs have been 

 in action since they were deposited, and the nodules must 

 consequently have been brought from a distance. There is 

 some difficulty however in accounting for the occurrence of 

 beds 4 feet thick, with no intermixture of foreign substances, 

 but consisting entirely of nodular limestone, on this sup- 



