4i8 



NATURE 



[August 23, 1906 



Italian Oisservations of the Total Solar Eclipse 

 ('905)- — An interesting illustrated report of the organisation, 

 equipment, and results of the Italian observations of the 

 total eclipse of August last is given by Prof. Ricco in 

 No. 7, vol. XX.KV., of the Memorie della Societa degli 

 Spettroscopisti Italiani. 



When first organised, the eclipse parly included Prof. 

 Tacchini, and, on his lamented death, the programme pro- 

 posed had, therefore, to be somewhat modified. 



Finally, it was decided that the expedition should make 

 its observations at AlcalS de Chivert, the programme 

 including spectroscopic and direct observations of the 

 prominences, photography of the corona, photographic 

 observations of the spectrum of the eclipsed sun with a 

 slit spectroscope and a prismatic camera, and observations 

 of the solar radiation, the ionisation of the atmosphere, 

 and the polarisation of the coronal radiations. 



.Although the work was interfered with by clouds, some 

 interesting and valuable observations were made, and are 

 recorded in the article referred to above. 



The Spectra of Sun-spots /nd Red Stars. — In a 

 previous paper Profs. Hale and .Adams considered the 

 question of the similarity of the spectra of sun-spots and 

 of fourth-type stars, and arrived at the conclusion that the 

 coincidences met with in comparing the spectra suggested 

 the existence of spots, similar to those on the sun, on such 

 stars. Their evidence was confirmed by Sir Norman 

 Lockyer, who further suggested that the temperature con- 

 ditions of fourth-type stars, taking the absorbing atmo- 

 spheres as a whole, are about the same as those obtaining 

 in the restricted region of a spot nucleus in the sun's 

 photosphere, both the stellar and the sun-spot atmospheres 

 having a lower temperature than that indicated by the 

 ordinary Fraunhoferic solar spectrum. In a paragraph 

 added to the present paper, reprinted as a Contribution 

 from the Solar Observatory, Mount Wilson (No. 8), Prof. 

 Hale acknowledges the possibility of this suggestion afford- 

 ing the true explanation. Prof. Hale's conclusion was 

 examined by Dr. W. M. Mitchell, who compared his 

 Princeton observations of spot spectra with the star spectra, 

 and was unable to confirm the coincidences of the 

 lines. 



In a paper now communicated to No. 5, vol. xxiii., of 

 the AstropJiysical Journal, Profs. Hale and Adams point 

 out that in the spectra of fourth-type stars the spot lines 

 may be obliterated by bright lines, and that their apparent 

 absence may not, therefore, be accepted as final evidence 

 until better photographs of the fourth-type spectra can be 

 obtained. Such spectra will probably be obtained when 

 the 5-feet reflecting telescope is erected at Mount Wilson 

 and a suitable spectrograph adapted to it. Comparing 

 the spot spectra with the spectrum of o Orionis, the same 

 observers show that the lines of the elements vanadium, 

 titanium, and manganese, which are strongly affected in 

 spot spectra, are also especially strong in this third-type 

 star. 



THE BRITISH ASSOCIATION. 

 SECTION G. 



ENGINEERING. 



Opening Address by J. A. Ewing, LL.D., F.R.S., 

 M.Inst.C.E., President of the Section. 



T intend to devote this Address to considering in certain 

 aspects the inner structure of metals and the manner in 

 which they yield under strain. It will not be disputed that 

 this is a primary concern of the engineer, who in all his 

 problems of design is confronted by the limitations imposed 

 on him by the strength and elasticity of the materials he 

 employs. It is a leading aim with him to secure lightness 

 and cheapness by giving to the parts such dimensions as 

 are no larger than will secure safety, and hence it is of the 

 first importance to know in each particular case how 

 high a stress may be applied without risk of rupture or 

 of permanent alteration in form. Again, the engineer 

 recognises the merit, for structural purposes, of plasticity 

 as well as strength, and in many of his operations he 

 NO. 192 1, VOL. 74] 



makes direct use of that property, as in the drawing of 

 wires and tubes or the flanging of plates. He is concerned, 

 too, with the hardening effect that occurs in such processes 

 when work is expended on permanently deforming a metal 

 in tlie cold state, and also with the restoration to the 

 normal condition of comparative softness which can be 

 brought about by annealing. Nor can he afford to be in- 

 different to the phenomena of " fatigue " in metals, which 

 manifest themselves when a piece is subjected to repeated 

 alternations or variations of stress — fatigue of strength and 

 fatigue of elasticity, which, like physiological fatigue, 

 admits under some conditions of rest-cure, inasmuch as it 

 tends to disappear with the lapse of time. No apology 

 need be made in selecting for a Presidential Address to 

 Section G a subject that touches so many points of direct 

 practical interest to engineers. It is a subject which has 

 for me the additional attraction of lying in the borderland 

 between engineering and physics — a borderland in which 

 I have often strayed, and still love to stray, and I enter 

 it to-day even at the risk of wandering into regions which, 

 to engineers, may seem a little remote from home, regions 

 where the landscape has, perhaps, a suspicious likeness to 

 that of the country over which the learned men of 

 Section A hold rule. 



To engineers, quite as much as to physicists and chemists, 

 we owe in recent years an immense extension of knowledge 

 regarding the structure of metals. This has come about 

 mainly by the intelligent use of the microscope. Take any 

 piece of metal, in the state in which an engineer makes 

 use of it, polish and lightly etch its surface, and examine 

 it under the microscope, and you find that it is a congeries 

 of a multitude of grains, every one of which may be proved 

 to be a crystal. It is true that the boundaries of each 

 grain have none of the characteristics of geometrica' 

 regularity which one is apt to look for in a crystal, but 

 the grain is a true crystal for all that. Its boundaries 

 have been determined by the accident of its growth in 

 relation to the simultaneous growth of neighbouring 

 grains — the grains have grown, crystal fashion, until they 

 have met, and the surface of meeting, whatever shape it 

 may happen to take, constitutes the boundary. But within 

 each grain there is the true crystalline characteristic — a 

 regular tactical formation of the little elements of which 

 the crystal is built up. It is as if little fairy children had 

 built the metal by piling brickbats in a nursery. Each 

 child starts wherever it happens to be, placing its first 

 brickbat at random, and then piling the others side by 

 side with the first in geometrical regularity of orientation 

 until the pile, or the branches it shoots out, meets the 

 advancing pile of a neighbour; and so the structure goes 

 on, until the whole space is entirely filled by a solid mass 

 containing as many grains as there have been nuclei from 

 which the growth began. 



We now know that this process of crystal growth occurs 

 not only in the solidification of a metal from the liquid 

 state, but in many cases during cooling through a 

 " critical " temperature when the metal is already solid. 

 We know also that the process may in certain conditions 

 go on slowly at very moderate temperatures. We know 

 also that the process of annealing is essentially the raising 

 of the metal to a temperature at which recrystallisation 

 may lake place, though the metal remains solid while this 

 internal rearrangement of its particles goes on. Whether 

 crystallisation occurs in solidifying from the liquid or 

 during the cooling of an already solid piece it results in the 

 formation of an aggregate of grains, each one of which is 

 a true crystal. Their size may be large or small — in 

 general, quick cooling means that crystallisation starts 

 from many nuclei, and the resulting grains are consequently 

 small ; with very slow cooling you get a gross structure 

 made up of grains of a much larger size. 



For simplicity of statement I shall ask you in what 

 follows to confine your attention to simple metals, omitting 

 any reference to alloys. Alloys present many complexities, 

 into which we need not at present enter. With simple 

 metals every crystalline grain is made of the same sub- 

 stance : the elementary brickbats are all exactly alike, 

 though there may be the widest variation from grain to 

 grain as regards the form of the grain, and also as regards 

 the direction in which the elementary brickbats are piled. 



