Till 



factory explanation of this remarkable phenomenon had been given. 

 The old theories, involving a supposed spheroidal concretionary tend- 

 ency in the material during consolidation, seemed quite untenable. 

 He examined with great care the appearances presented in the sur- 

 faces of the stones, and concluded that the columnar structure is due 

 to the shrinking and cracking during cooling of a very homogeneous 

 mass of material. The cross joints he considered to be in reality 

 circular conchoidal fractures commencing at the centre of the column 

 and flashing out to the circumference. 



A very interesting subject, and one of very high importance, to 

 which Professor Thomson gave great attention, is the flow of water 

 in rivers. He investigated, with great care, and from a theoretical 

 point of view, the origin of windings of rivers in alluvial plains, and 

 his conclusions were published in the 'Proceedings of the Royal 

 Society,' May 4, 1876. Later in the same year he constructed, in 

 clay, on a table, a model with which he investigated the movements 

 of the different parts of the water in passing round the bends in this* 

 artificial river ; and, finally, he made a large wooden model of a river 

 flowing on a nearly horizontal bed with many bends and various 

 obstacles. By aid of fine threads, small floating and sinking bodies, 

 and coloured streams of fluid coming from particles of solid aniline 

 dye dropped into the channel, he was able to follow from point to 

 point the movements of the fluid, and thus to give not only beautiful 

 and striking ocular evidence of the truth of his early conclusions, but 

 also to extend his theory. Papers on this subject were communicated 

 to the Royal Society in 1876, 1877, 1878. The paper of the last- 

 named date was entitled " On Flow of Water in Uniform Regime in 

 Rivers and in Open Channels generally." It contains a very clear 

 and striking account of what does occur in the motion of a river 

 down its inclined channel ; and, in particular, of the fact which seems 

 to be ascertained, that the forward velocity of the water in rivers is, 

 generally, not greatest at the surface with gradual abatement from 

 surface to bottom (as would be required under the conditions sup- 

 posed in the laminar theory) ; but that, in reality, the average velocity 

 down stream is greatest at some depth below the surface, from which, 

 up to the surface, there is a considerable decrease, and down to the 

 bottom a much greater decrease. This phenomenon he showed very 

 clearly to be due to the rising, of masses of slow-going water, from 

 the bottom, on account of directing action of bottom obstacles. These 

 masses of slow-going water, when they reach the top, spread them- 

 selves out, and, mingling with the quicker surface water, give to it, 

 on the whole, a less rapid movement than it should otherwise possess. 

 The paper, as a whole, forms a masterly exposition of this important 

 subject. 



Finally, in this brief summary must be mentioned the paper which 



