142 Correspondence — Rev. Dr. Irving. 



of the upper layers can neither help {per se) the downward flow 

 when the terapei'ature rises again, because the expansion being 

 equal to the previous contraction (cet. par.), it must result merely 

 in the closing-up of the shrinkage-cracks ; nor can it tend to produce 

 a curvature of the lower layers (after the fashion of the balance- 

 wheel of a chronometer) as seems to be suggested later on in Mr. 

 Goodchild's paper. In the former case water, produced by the 

 melting either of the surface of the glacier or of the snow-mantle 

 overlying it, during hours of sunshine, and flowing into the cracks, 

 would certainly by its expansion in freezing do some work ; but 

 what becomes of its latent heatf The distribution of this in the 

 neighbouring ice needs to be considered. 



2. If by ' cold-waves ' Mr. Goodchild means (as I take it) ' flows ' 

 of heat by conduction from the warmer interior to the surface 

 whose temperature is below 0° C, it is a pity he did not speak of it 

 as such. Cold, like darkness, is a negation : and we can only speak 

 of a wave of either metaphorically. 



3. We had no need to go back nearly half a century to Brunner's 

 investigations to convince us that the ' sole ' of a glacier moves down 

 a slope ; the observations and measurements by Tyndall, Helmholtz, 

 Forbes, Agassiz, and others have made that pretty certain. 



4. As to the "uphill movements" postulated by Mr. Goodchild 

 and many writers who have preceded him, I have for a long time 

 been very sceptical, as may be seen from my papers referred to 

 above ; and the researches of Penck ^ on the glaciation of the 

 Northern Alps have conveited that scepticism into positive disbelief, 

 since all that I have read or heard alleged, as evidence of such 

 movements, is more rationally explained by the overflow of the 

 glaciers beyond their valley-sides during periods of maximum 

 glaciation, and by the phenomena of stranded lateral moraines 

 during the recession of a glacier, with which every Alpine observer 

 is familiar (e.g. the Morteratsch) . 



5. The theory of " isogeothei-ms " continued through the lower 

 parts of a glacier will not work, except for a hypothetical case 

 where the adjacent rocks and the ice were all considerably below 

 0° C, because no heat at any higher temperature could be conducted 

 through or into the ice, as seems to be imagined : it would become 

 latent in the melting of the ice at the contact. This is probably 

 the reason why previous writers have " overlooked " the fiction 

 which Mr. Goodchild has now gravely put forward as a " fact." 



6. Mr. Goodchild has overlooked three factors essential to the 

 construction of any sound physical theory of glacier-motion ; (i.) 

 liquefaction under pressure and regelation; (ii.) the "greenhouse- 

 principle," the application of which to glacier-motion was demon- 

 strated by myself in the paper in Nature already referred to ; 

 (iii.) the latent heat of water, the 80 gramme-units of heat given up 

 by every gramme of water at 0° C. in the act of solidifying under 

 ordinary pressure. A. Irving. 



Wellington College, Berks, Jan. 8. 

 1 See " Die Vergletscherung der Deutschen Alpen," reviewed in the Geol. Mag. 

 Dec. II. Vol. X. p. 174, et seq. 



