320 



NA TURE 



[February 3, 1898 



the sanitary aspects of the beds are noticed. In the 

 map the index of colours is grouped in three series, 

 clayey, gravelly and sandy, with Alluvium detached at 

 the top, and Chalk at the bottom. 



In the rest of the memoir a purely sanitary arrange- 

 ment is adopted, under four heads, the first being that of 

 the sub-soil with reference to sites for houses (pp. 27-32). 

 The way in which not only the character of the beds, but 

 also their thickness, position, and mutual relations affect 

 the suitability of ground for building-purposes is enforced, 

 and the causes of contamination of porous sub-soils are 

 discussed : even gardens are not neglected. 



Secondly, water-supply and drainage are treated (pp. | 

 33-39). After noticing the supply of London, the question | 

 of rural water-supply is taken, the causes of contamina- 

 tion of shallow wells and the danger of "dead wells,' 

 used as receptacles for sewage, &c., being described. 

 Then we have sanitary considerations in regard to the 

 situation and surroundings of houses (pp. 40-45)) under 

 which head subjects other than geologic are referred to, 

 such as surface-drainage, fog, sunshine, rain, wind, floods. 



The troublesome question of cemeteries fittingly comes 

 last (pp. 46-48), and the author concludes that "an 

 isolated tract of elevated ground, where sands and sandy 

 loams, or sandy and loamy gravel, of considerable thick- 

 ness, rest on clay also of considerable thickness, offers 

 the most desirable site," which, however, is qualified by 

 the addition that '' probably a sandy and calcareous 

 loam is the best material for a graveyard." However, 

 the difficulty is to get sucTi desirable sites ! The fitting 

 conclusion is a list of the cemeteries in and near London, 

 with the beds on which they are placed, by means of 

 which people who are thoughtful of their latter end can 

 select the geologic formation in which they would like to 

 be buried. 



The long index (ten pages), is really more than an 

 index. It is not limited to giving the pages of reference ; 

 but also notes, by figures in another type, the height 

 of the various places above Ordnance datum, and, by 

 means of letters, the beds on which the places are— an 

 ingenious and useful novelty. 



Although this is the first Geological Survey memoir 

 devoted to sanitary matters, it should be remembered 

 that stores of information that is useful from a sanitary 

 point of view are to be found in many of those memoirs, 

 especially as regards wells and water-supply. 



This work should have a ready sale, as it is written in 

 such a way as to be useful to the ordinary inquirer, and 

 of interest to any intelligent reader ; but the price {zs. bd.) 

 is rather high, presumably on account of the map. It is 

 to be hoped that the author may ere long have to prepare 

 a new and enlarged edition, and that the success of the 

 present venture may lead to other work of the same sort 

 being undertaken. W. Whitaker. 



PROFS. C. RUNGE AND F. PASC HEN'S 



RESEARCHES ON THE SPECTRA OF OXYGEN, 



SULPHUR, AND SELENIUM} 



T N the above paper Profs. Runge and Paschen have 

 ^ extended their important investigations to the spectra 

 of oxygen, sulphur and selenium, and have discovered 

 in the low temperature spectra of these elements a 

 number of series which are very similar to those 

 previously found for other elements. The principal 

 characteristics of these series are too well known to need 

 any further explanation ; they may be represented by 

 Kayser and Runge's original formula, 



n'- n* 



where N is the wave frequency and n takes up the values 

 3, 4, 5 . . . for the different lines of the same series. In 



1 IViede.iianns AiinaUn, vol. Ixi. p. 641. 



NO. 1475. VOL. 57] 



the present cases n^ may be with advantage replaced by 

 n^. We may with equal accuracy take Rydberg's 

 formula, 



N = A - B(« + fi)-^ 



where A, B, /i are the three constants. In the spectrum 

 of oxygen which goes by the name of " compound line 

 spectrum," and which Piazzi Smyth has shown to be 

 chiefly made up of close triplets, Runge and Paschen 

 show that these triplets arrange themselves in two series, 

 having approximately the same convergence frequency. 

 The difference in the curve numbers of the components 

 being the same for each triplet, they have all the charac- 

 teristics of the " Nebenserien " or " associated series," as I 

 propose to call them, because they always occur in 

 groups of two. The formulas for the strongest lines of 

 the triplets in Rydberg's form is 



1st assoc. series : 23207, 96-1 10396 (« - o"02i48)"- 

 2nd ,, „ 23200, 63 - 10901 1 ^w- 0-24127)"- 



Rydberg has given a law according to which the 

 principal series may be calculated from the associated 

 series, with sufficient approximation to identify the lines 

 belonging to it in the present instance. Only two triplets 

 belonging to it have been found ; the first in the extreme 

 red has a wave-length of 7776 for its strongest lines, and 

 the second, photographed by Runge and Paschen for the 

 first time, has a wave-length 3948. In addition to the 

 triplets, the authors have discovered a number of 

 doublets which can also be grouped into two associated 

 series, having the equations 



21205, 56- 109366, 7 (« -o'i6i9i)"^ 

 21211, 11-110346, 7 («- 001093)"'^ 



and the principal series belonging to this group is 

 probably represented by two lines at wave-lengths 4368, 5, 

 and 3692, 6, the former being one of the strongest lines 

 of the compound line spectrum. 



The result of this investigation of the oxygen spectrum 

 is, that it shows two sets of associated series similar to 

 that found in the case of helium, and that therefore the 

 spectroscopic evidence that helium is a mixture of two 

 gases no longer holds. There is very little doubt that 

 the oxygen spectrum is represented among the Fraun- 

 hofer lines ; almost conclusive evidence in favour of this 

 being given by the first triplet of the principal series, 

 which fells at 7776 in a portion of the solar spectrum 

 which is comparatively free from lines. The beautiful 

 photographs of Higg show at this place a triplet the 

 components of which have exactly the right difference in 

 frequency. 



Profs. Runge and Paschen also investigated the spectra 

 of sulphur and selenium, and discovered spectra which 

 correspond to the compound line spectrum of oxygen. 

 The spectra consist chiefly of triplets which may be 

 arranged in a group of associated series ; and there are 

 also indications of the existence of two principal series in 

 each case. The authors apparently considered these 

 spectra as due to the elements, but they have only been 

 able to obtain them in the presence of oxygen. Further 

 investigation is therefore required to show that they are 

 not really oxide spectra. Should this prove to be the 

 case it would be a matter of some importance and 

 great interest, for it would show that we must consider 

 the compound line spectrum of oxygen as due to a corn- 

 pound molecule, an " oxide of oxygen " similar in consti- 

 tution to the oxides of sulphur and selenium which give 

 the analogous spectra. 



In conclusion, I may add a few remarks as to the rela- 

 tive merits of Rydberg's and Kayser and Runge's equa- 

 tions. The greater simplicity in form of 



N = A - ^,-?, 



adopted by Kayser and Runge, and the ease with which 

 the constants may be calculated would, independently of 



