July i, 1920] 



NATURE 



563 



dry, these cannot be lumped together for correlation 

 purposes, as the whole effect will be masked. We are 

 reminded of the sun-spot maximum of 1893, which 

 was associated with great heat in England and 

 France, but was exceptionally cold in America and 

 other parts of the world. This limitation of districts 

 may not, as the author recognises, be the same for 

 short periods as for long ones, but he finds the major 

 characteristics in mountain regions very much alike 

 over distances of fifty or sixty miles, and relies upon 

 the evidence of the trees themselves for the demarca- 

 tion of the districts. 



One other small difficulty Prof. Douglass has met 

 in an ingenious manner. It is often noticed that such 

 an element as rainfall, when expressed as departure 

 from the mean, as it must be in correlation problems, 

 is arithmetically lacking in symmetry, since the 

 defect can only be 100 per cent, at most, while excess 

 can be very much larger. Geometrically, this can be 

 avoided by using a logarithmic scale, but this flattens 

 the variation very much. Prof. Douglass's device is 

 to leave the deficient amounts unaltered, but in the 

 case of excessive falls to invert the fraction and 

 measure upwards from the normal. Thus a rainfall 

 of twice the normal is indicated by a ooint just so far 

 above the normal line as the point indicating a rainfall 

 of half the normal is below it. The symmetry is not 

 perfect, as, of course, no possible wetness can give a 

 point corresponding to zero rainfall, but the method 

 is convenient in places where zero rainfall in the unit 

 period is unknown. W. W. B. 



The Interferometer in Physical Measure- 

 ments.^ 



A FOURTH volume describing the researches of 

 Prof. Carl Barus with interferometers has 

 recently been issued. The classical work of Fizeau, 

 who applied interference methods to the determina- 

 tion of expansion coefficients, directed attention many 

 years ago to the possibility of the kind of work which 

 has been so well developed by Michelson and others, 

 and in the present series of papers Prof. Barus seeks 

 to develop the methods of application of the interfero- 

 meter to a somewhat wide range of physical measure- 

 ments. These include spherometer measurements, 

 elastic deformation of small bodies, elongations due 

 to magnetisation, pressure variation of specific heat 

 of liquids, and even electrodynamometry. The re- 

 mainder of the volume deals with various modifica- 

 tions of the interferometer methods and with certain 

 gravitational experiments. 



Doubtless such an investigation of methods will 

 be useful to workers in any of the foregoing fields, but 

 so far as a first impression is to be trusted it would 

 appear that the main interest has lain in the method 

 rather than in any results which have been attained. 



In order to study the motion of a contact lever, it 

 may be made to carry two small mirrors reflecting 

 normally two beams which are afterwards caused to 

 interfere. Any rotation of the lever obviously causes 

 a difference of path, which appears in the shifting 

 of the easily recognisable and distinctive central 

 " achromatic " interferometer fringes, such motion 

 being measured by a plate micrometer or "graticule" 

 in the observing telescope. 



The two mirrors form the limbs of a "T" piece, 

 which is pivoted about a hinge at the end of the foot. 

 One limb ends in a contact pin which abuts against 

 the surface, the motion of which is to be measured. 



In such circumstances Prof. Barus estimates the 



1 " Disp'acenri'nt Intfrferometrv by the Aid of the AcViromat-c Frinjtes." 

 Part iv. r.y Prof. Carl Baru«. (Carnegie Institution of Washington, 1919.) 



NO. 2644, VOL. 105] 



limiting sensitiveness to be 33x10-° cm., or perhaps 

 even a third of this amount, but it should not be 

 forgotten that the very simple interferometer system 

 of an optical test-plate has a sensitiveness of about 

 a quarter wave-length, say 12x10-* cm., and this 

 without a doubtful hinge and another contact. The 

 contact lever can, of course, deal with non-specular 

 surfaces, but to use it as a spherometer for a glass 

 lens seems quite needless. Naturally, an apparatus 

 of this nature is excellently adapted to such a problem 

 as that of investigating the changes of length of a 

 magnetised rod, and, although no very novel results 

 are obtained, the investigation has been compara- 

 tively easy, and the method is well adapted for 

 demonstration. 



Suitable self-adjusting interferometers, such as are 

 described in chap, vii., ought to find an increasingly 

 useful place in the physical laboratory, and students 

 should be taught the practical use of such instruments 

 and their modifications. There is too great a tendency 

 to treat an interferometer as a piece of apparatus 

 sacred to one or two highly specialised purposes, but 

 with little more than a few pieces of good plane 

 parallel glass a set of instruments can be made up 

 which should be of the greatest use in teaching and 

 research. 



One could wish, perhaps, that some one problem 

 had been attacked and solved thoroughlv. The 

 curiously unfinished nature of the work is disappoint- 

 ing, but we must conclude that the method is the 

 chief object. As regards the text, the descriptions 

 are clear and praiseworthy, but the diagrams are 

 both inadequate and unsatisfacttDrv. L. C. M. 



Canvas-destroying Fungi. 



AIT HEN men again began to take to their tents at 

 *' the outbreak of war, many noticed that dark 

 brown and black spots, frequently of a diamond shape, 

 were not uncommon on the canvas. Small, sur- 

 reptitiously acquired bits began to be scattered around 

 for information as to the identity of the moulds caus- 

 ing the rot. Now it is very surprising that so little 

 work has been done on canvas-destroying fungi. That 

 canvas is liable to suffer from moulding seems generally 

 to be known, judging from the fact that any 

 material likely to get wetted is usually "cutched." 

 Shortly before the war aircraft workers began to 

 interest themselves in the fungi concerned in the 

 damage, but it was not until war broke out that 

 one realised the extent of the destruction of sails, 

 tents, etc., by these organisms. 



Major W. Broughton-Alcock, in the Journal of the 

 Royal Army Medical Corps for December last, gives 

 a short account of investigations carried out by him 

 in Malta, Italy, and (in conjunction with Miss A. 

 Lorrain Smith) at the Natural History Museum. In 

 Malta attention was soon attracted to the rapid 

 spotting and destruction of tentage — awnings last there 

 only about a year. The investigators found that the 

 principal agents of destruction of cotton- and flax- 

 made canvas are Macrosporium and Stemphylium. The 

 latter is the more prevalent in Malta, and could be 

 isolated by exposing culture plates to the air. The 

 colours of the spots on canvas correspond to the 

 colours seen in cultures, being first brown and then 

 black. The variation in the colour of the spots, 

 especially noticed in flax-made and more resistant 

 canvas, was found to be due to other fungi in asso- 

 ciation with the above genera — Septoria, Alternaria, 

 Helminthosporium, Chaetomium, Exosporium, Peni- 

 cillium, Oospora, Torula, Saccharomyces, and yellow 

 pigment-forming and other air-borne bacteria. Though 



