6o 



NA TURE 



[November i6, 189; 



is just twice as long as l, the fringes will appear simul- 

 taneously on both upper surfaces, D and B. 



The adjustment of the length of the standards is usually 

 made to within a few waves, and the outstanding differ- 

 ence is measured by a compensating device. 



This is furnished by the rotation of the compensating 

 plate, Go, Fig. 8. The plate is held in a metal frame 

 which is supported at one end by a short thick rod firmly 

 fixed to the bed. At the other end a delicate spiral spring 

 is attached ; the tension of the spring tiuists the rod 

 through a minute angle, and thus alters the thickness of 

 glass traversed by one of the interfering pencils. The 

 other end of the spring is attached to a flexible cord pass- 

 ing over a pulley which is connected with a graduated 

 circle. The angular motion is thus reduced about 100,000 

 times, and yet the proportionality is preserved. 



Suppose the outstanding .difference is f a fraction of 

 a wave-length known to within one or two tenths, then 



II = 21 + e 



and consequently the number of red waves should be 

 2 X I2i2'34 + f. This fraction is corrected by direct 

 observation, as in the case of standard I, and the same 

 control is furnished by the concordance of the results for 

 the three colours ; so that an error in the whole number 

 of waves is well-nigh impossible. 



The process of comparison and correction is repeated 

 in the same way with the other standards, until we finally 

 arrive at the whole number of waves and approximate 

 fraction in the 10 centimetre standard. Up to this point 

 the question of temperature and pressure is of minor 

 importance, for the comparisons and corrections are 

 made while both standards are under the same con- 

 ditions ; and being all made of the same material, it is 

 sufficient to know that the temperature is the same for 

 both. In the measurement of the fractions on the 10 

 centimetre standard, however, it is necessary to know 

 the temperature and pressure with all possible accuracy, 

 and it is also important that the comparison of this 

 standard with the metre should be made, as nearly as 

 may be, under the same conditions as that of the deter- 

 mination of the standard in light-waves. 



The author having been honoured by an invitation 

 from the International Bureau of Weights and Measures 

 to undertake a series of experiments upon the lines here 

 briefly indicated, the necessary apparatus was constructed 

 in America, and shortly afterward installed in the Bureau 

 International des Poids et Mesures at Sevres. 



Two complete and entirely independent determinations 

 were made. These have not yet been completely re- 

 duced, but an approximate calculation gives for the 

 number of waves of red light in one metre of air at 

 15' C. and 76 mm. 



1st series ... ... ... ... I553i63'6 



2nd series I553i64'6 



The difference from the mean is half a wave, or about 

 one fourth of a micron.^ 



From these results it follows that we have at hand a 

 means of comparing the fundamental standard of length 

 with a natural unit — the length of a light-wave — with 

 about the same order of accuracy as is at present possible 

 in the comparison of two metre bars. 



This unit depends only on the properties of the 

 vibrating atoms of the radiating substance, and of the 

 luminiferous ether, and is probably one of the least 

 changeable quantities in the material universe. 



If, therefore, the metre and all its copies were lost or 

 destroyed, they could be replaced by new ones, which 

 would not differ from the originals more than do these 

 among themselves. While such a simultaneous destruc- 

 tion is practically impossible, it is by no means sure that, 



1 The error in the determination of the relafi'z'e wave-lengths of the three 

 radiations is very much smaller, probably less than one twenty-millionth. 



NO. 1255. VOL. 49] 



notwithstanding all the elaborate precautions which have 

 been taken to insure permanency, there may not be slow 

 molecular changes going on in all the standards ; changes 

 which it would be impossible to detect except by some 

 such method as that which is here presented 



A. A. MiCHELSON. 



FURTHER XOTES AXD OBSERVATIONS 

 UPON THE INSTINCTS OF SOME COMMON 

 ENGLISH SPIDERS. 



ly/r ANY of what would otherwise be most interesting 

 ^*- anecdotes respecting the habits of spiders have 

 been related by persons who, being unacquainted with 

 the immense number of " kinds" of this group that there 

 are in England, not to mention the rest of the worlds 

 have apparently considered that all needful information 

 in the way of the animal's identity has been supplied by 

 the simple statement that it is a spider. 



Such anecdotes have of course a certain value, inas- 

 much as they furnish some general information respecting 

 the instincts of the class as a whole. But to those who 

 are anxious to compare together the instincts of indi- 

 viduals of the same or different species, genera, and 

 families, who are anxious to acquire in short some little 

 knowledge of the comparative psychology of the group, 

 they are distressingly incomplete. 



To remedy in part these deficiencies, to verify the 

 experiments of others, and to make fresh observations 

 upon some points that are open to dispute, I took the 

 opportunity, during a recent visit to North Cornwall, of 

 compiling a set of notes upon the habits of some of the 

 commonest spiders in the neighbourhood. 



In the following paper, which is based upon these 

 notes, I have added some brief accounts of the webs, 

 habitats, or general appearance of the spiders, so that 

 those persons who are not acquainted with the animal by 

 name, may yet, with but little trouble, ascertain what the 

 species are that are under discussion. 



Agalcna labyrijithica. — This spider may be looked 

 upon as the country cousin of the common house spider, 

 Tegcjian'a atrica, which being essentially a lover of 

 bricks and mortar, is found in lofts, disused rooms, &c., 

 where it spins in corners and other angles a horizontal, 

 triangular sheet of web, a familiar structure which must 

 be associated in all minds with the word cobweb. 



The snares of Agalena are essentially like those of 

 Tegenarid, consisting of a short silken tube or funnel, 

 one end of which is buried in the bush that the spider 

 has chosen to build in, while the other opens upon, and 

 is continuous with, a widely extended horizontal sheet 

 composed of fine closely woven silken threads. During 

 the daytime the spider, if cautiously approached, may 

 usually be seen squatting at the entrance of her funnel. 

 She is, however, remarkably wary, and this, coupled with 

 her equally remarkable agility, makes the task of 

 capturing her by no means an easy one. For, by means 

 of the further open e.xtremity of the tube, she can make 

 her escape into the bush beyond. Wherever I have had 

 an opportunity of observing this spider, I have noticed 

 that it appears to have a special liking for furze bushes ; 

 and it seems reasonable to suppose that this selection of 

 so prickly a site for the building saves the young and 

 also the nest from destruction at the hands, or rather the 

 noses and legs, of cattle. 



Upon examining the debris of prey, with which the 

 orifice of the funnel was usually strewn, I was surprised 

 to find that it consisted more often of the remains of 

 bees than of flies — generally, indeed, the limbs, wings, 

 <S;c., were those of some species of Bumble-bees iBombiis). 

 Being curious to see how the spider would manreuvre to 

 overcome so redoubtable an adversary, I captured upon 

 one occasion a small specimen of a bumble-bee and 



