530 



NATURE 



\_April I, 1884 



length of the metre has been marked by means of two 

 fine lines. The position of the lines at a constant tem- 

 perature is then determined by the micrometers, the bar 

 being placed for this purpose in a trough of water, the 

 temperature of which is maintained constant by an im- 

 proved automatic regulator. A second metal bar, whose 

 rate of expansion is to be determined, is placed in a 

 separate trough of water, the temperature of which differs 

 considerably (rem that in the other trough. This trough 

 is then also brought into position under the microscopes, 

 and the positions of the lines on the second bar deter- 

 mined relatively to those on the first bar. This method 

 has the advantage that the results are independent of any 

 change in the distance between the axes of the two 

 microscopes during the comparison of the two bars. The 

 optical eflect of the immersion of the bars in water was 

 investigated by M. Krusper in 1872-73, who found it to 

 affect the comparisons very little. 



The comparing apparatus at the Bureau was originally 

 made by M. Sorensen of Stockholm, but was subse- 

 quently altered and improved by the Geneva Society for 

 the construction of physical instruments, under the direc- 

 tions of M. Turettini. The lines on the bars were 

 illuminated by light reflected on to a small mirror fixed 

 at an angle of 45^ inside the microscope, a little above 

 the object glass. The determinations of the errors of 

 each micrometer-screw throughout its whole length, for 

 even no micrometer-screw has yet been made in which 

 appreciable errors may not be detected in its use, was 

 made in accordance with methods followed by Drs. 

 Foerster and Hirsch, and MM. Starke and Kammerer. 



The thermometers used were constructed after the 

 form adopted by the Bureau (tome i. p. B 8), and were 

 made at Paris by M. M. Baudin and M. M. Alvergnat. 

 It is satisfactory to find that to the important question of 

 thermometers the Bureau has given much attention, as in 

 such investigations errors of thermometers are of as great 

 importance as the errors of the micrometer-microscopes, 

 but are not, however, always so carefully attended to as 

 they should be. The thermometers were calibrated after 

 the methods suggested by Dr. Thiesen and M. J. Marek 

 (" Repertorium der Carl," t. xv. 1879), and were corrected 

 for "exterior pressure" toa barometric height of 760mm. 

 at 0° lat. = 45°, as well as for " interior pressure," or 

 vertical position, the thermometers reading from o^-o2 to 

 o°o6 C. too high when placed in a horizontal position. 



During the past years this apparatus has been used in 

 determining the rates of expansion of the platinum-iridium 

 metres deposited at the Bureau, which are intended here- 

 after to be the universal standards or prototypes of the 

 metric system. The linear coefficient of expansion for 

 1° C. of the platinum-iridium was found to vary from 

 o'ooooo866S to o'ooooo8689, with a probable error of only 

 + 00000000075. 



The high accuracy of the results obtained at the Bureau 

 in the weighings there executed, have been already pre- 

 viously referred to, as they appeared in a separate form 

 in 1881. In the present volume M. Marek gives the 

 particulars of the experiments made by him in redeter- 

 mining the density of mercury of the kind actually used 

 in barometer tubes, taking the mean density of mercury 

 as being comprised between that of perfectly dry mer- 

 cury and of mercury exposed to moist air. Illustrations 

 are given of the modes of purif\ing and of weighing the 

 mercury. The density of four samples of mercury, as 

 determined by weighing in water, was found after many 

 experiments to be as follows : — 



Mercury A = I3'59563i ± 0*000029 



H = I3'S95633 ± 0-000024 



,, C = 13-595458 ± o'cooc56 



D = 13 59593° =t 0-000055 



In the paper, " Dilatation du Mercure," we find again 

 that painstaking investigat'on and high accuracy which 



characterised the papers published in 1881 above referred 

 to. The most exact observations on the dilatation of 

 mercury are undoubtedly those of M. Regnault {JM/nwires 

 de P Academic dcs Sciences, tome xxi. 1847); and it is 

 to the mathematical reduction of these observations that 

 Dr. Broch has now applied a critical examination, em- 

 ploying as his first coefficient of dilatation the value 

 obtained by M. WuUner (" Lehrbuch der Experimental 

 Physik," t. iii.): — 



(/, = io-» (1S1168 \- 11-554^ -f 0-021187/=), 

 instead of that of Regnault— 



d, = 10-" (179007 -f 25-232/). 

 By a reduction by the precise method of least squares, 

 of the original observations to the latitude of 45° at the 

 level of the sea {B = 760 mm.), there is now obtained 

 for the cubic expansion of mercury the following formula, 

 which we would recommend to the attention of those 

 engaged in accurate work : — 



I +/{■/ = I -x- o'oooi8i792 . / -^ 0-000,000,000175 . C- -V 

 0-000,000,0000351 16 . /■'. 



We note that for the current year the President of the 

 Bureau is General Ibanez (Madrid), the Secretary being 

 Dr. Hirsch (Neuchatel), the Committee including MM. 

 Dumas (Paris), Foerster (Berlin), Gould (Cordoba), Govi 

 (Naples), Herr (Vienna), Hilgard (Washington), Krusper 

 (Budapest), Stas (Brussels), Wild (St. Petersburg), and 

 Wrede (Stockhohn). This country is not represented on 

 the Committee, our Government having decided not to 

 take part in this international project. 



LIL.EA 1 



THE genus Lilaa was founded by Humboldt and 

 Bompland for a very curious plant closely allied to 

 our native Triglochin, which was first found by them in 

 New Grenada. The present memoir, which has ap- 

 parently only recently reached Europe, is one of the most 

 elaborate studies probably ever made of the entire mor- 

 phology, histology, and development of a single flowering 

 plant, and is due to the unexpected discovery of the 

 plant in 1875 in the Argentine Rtpublic. The curious 

 reductions of structure which are the result of a more or 

 less aquatic mode of life have always made plants of this 

 kind attractive to investigators. 



The careful investigation of the structure of the flower 

 throws some light on a point w'hich has been much con- 

 troverted, whether the stamen is ever an axial structure 

 or not. Lilcca bears its flowers in a spike, and there are 

 no less than three kinds : —(i) below, female ; (2) in the 

 middle, hermaphrodite ; (3) at the top, male flowers. 

 These latter consist of a single stamen in apparent direct 

 prolongation of the floral axis. It is about these in the 

 similar cases of Naias that discussion has arisen. Now 

 Hieronymus contends that this stamen is really only 

 pseudo-terminal, but that it consumes in its development 

 the primitive meristem of the growing point, and so 

 eventually occupies its place. He extends the same 

 explanation to the cases of Naias, Zaniiiclicllid, Casua- 

 riiia, Briziila, and others which have been held to sup- 

 port the axial origin of stamens. But as Sachs remarks 

 ("Textbook,'' second edition, p. 541), the question cannot 

 be settled wholly on anatomical grounds. And in Lilcea 

 there can be no doubt that in the hermaphrodite flowers 

 the stamens are lateral. In the male flowers he some- 

 times finds a lateral rudiment of a pistil; and this must be 

 held to clinch the argumei t that the stamen is not really 

 cauline, but always lateral and only pseudoterminal. 



Lilcca has a fourth class of flowers, the adaptive origin 

 of which is interesting. The whole plant is at first par- 

 tially submerged — perhaps was once wholly so. The 



* "Monografia de Lilaea Fubulata." Por J. Hieronymus. Ac'as lic In 

 ni lie Cieiicias en C6rdoba. (Ruenos Aires, 1882.) 



