903 



NATURE 



[December 22. 1923 



On one occasion when developing a photographic 



})late at the Observatory I received a slight shock 

 rom a lightning discharge nearby, or more probably 

 from an induced charge in the lead lining of the 

 developing table. Since then I have hesitated to 

 go up to the Observatory during thunderstorms, 

 and have been accustomed to look out for the " all 

 dear " signal which these mammato-cumulus clouds 

 give us. 



Another remarkable fact connected with local 

 thunderstorms at Kodaikanal, and probably else- 

 where, is the curious roaring soun<l emanating from 

 the cloud before a storm begins. At first 1 con- 

 sidered this was due to heavy rain approaching, but 

 concluded that this could not be so. The situation 

 at Kodaikanal is such that one may find oneself very 

 near to a cloud mass rising over the steep sides of 

 the mountains, and the sound always appears to 

 come from the cloud itself, and not from the ground 

 or from trees. Possibly Dr. Simpson can explain 

 this ; he would probably have heard it at Simla. 



J. EVERSHED. 



Ewhurst, Surrey, 

 November 25. 



Consumption of Fish by Porpoises. 



In the course of our cruises, we have often har- 

 pooned porpoises of various species, and occasionally 

 mvestigatecl the contents of the stomach. Some- 

 times the stomach was found to be empty, but in 

 most cases it contained remains of fish, though 

 these were, as a rule, so decomposed by the digest- 

 ive fluids that identification was impossible. Now 

 and again, however, it could be done. In the 

 vicinity of the Continental Slope, for example, west 

 of the English Channel, where porpoises are nearly 

 always found in abundance, we found great bundles 

 of the pelagic pipe-fish, Entelurus (squoreus L., in 

 the stomachs of porpoises taken. But as a rule, 

 the porpoise evidently prefers fish of somewhat 

 more fleshy build than the pipe-fish. 



The present note is occasioned by the recent pre- 

 liminary investigation of a sample from one of the 

 cruises of the Thor. On June 24, 1910, being then 

 off the south coast of Spain, in the Mediterranean 

 (between 36° 10' N., 4° 42' W., and 36° 19' N., 

 4° 0-6' W.), we harp>ooned a female specimen of the 

 common long-nosed porpoise {Delphinus delphis L.). 

 The stomach contents consisted of fish-residue : 

 more or less dissolved soft parts, crumbling back- 

 bones, otoliths, and eye-lenses. I noted that most of 

 the fish-bones were green, but no identification was 

 attempted. The most interesting feature was the 

 great number of otoliths, or ear-bones, of fish. When 

 these were sorted out and counted, there were no 

 fewer than 15,191 of different sizes, though mostly 

 small. Several species were represented, about five ; 

 but, owing to lack of material for comparison, I can- 

 not give any further determination at present. 

 Some are presumably those of Scomberesox, clupeoids 

 and scomberoids, possibly also scopelids. 



The sample is interesting, inasmuch as it gives 

 some slight idea of the p>orpoise's enormous con- 

 sumption of fish : in the stomach of this one specimen 

 we found remains of no fewer than 7596 fish. How 

 long the porpoise took to collect the whole 7596 

 it is impossible to say, since we do not know how 

 long the otoliths remain in the stomach before being 

 dissolved or passed out. The fact that otoliths are 

 not always found in the stomachs of porpoises seems 

 rather to suggest that they do not remain there 

 very long. 



JoHs. Schmidt. 



NO. 2825, VOL. 112] 



Crystallisation of Cementite in Steel. 



With reference to the particularly interest m; 

 article by H. C. H. C, in Natuke of November 1; 

 p. 728, might I mention the following amongst matI^ 

 other examples which have come under my notw ■ 

 illustrating the tendency of cementite to form c: 

 walls or a network under con<Iitions where the occu- 

 rence of poarlilc is more commonly anticipate<l ' I 



Fi(j. I. — Cementite uetwork in mild ktr 



dead mild steels the occurrence of cementite m either 

 network or comparatively massive formation has been 

 recognised by a number of investigators. Fig. i 

 illustrates an exceptional case in which isolated tell 

 walls were found only near the edge of a dead mild- 

 steel plate, in a region otherwise microscopically 

 carbonless. The only apparent explanation of this 

 occurrence was that the plate must have become 

 carburised locally during the processes of manufacture. 



1 .■-.. z. Ccinciuite in ferrite grain junctions, in nickel steel, v 1500. 



In the alloy steels, the simultaneous occurrence of 

 structurally free ferrite and juxtaposed carbide net- 

 work would be regarded as uncommon, but Fig. z 

 shows an instance of this occurrence in a large oil- 

 hardened nickel-steel forging. At the magnification 

 of 2500 with which the original photograph was taken 

 by means of the super-microscope, the carbon-con- 

 taining constituent can readily be seen to be lamellar 

 pearlite, which makes easy the identification of the 

 carbide plates in the ferrite boundaries. 



F. Rogers. 



64 A Westbar, Sheffield. 



