55S 



NA TURE 



[October 5, 1905 



points of maximum absorption, or most persistent edges 

 in each of the bands measured by Hartley and Dobbie. 

 This is shown in the following statement quoted from his 

 paper (Wilhelm Friederichs, Zcit. fiir nnssenschapliche 

 Photographic, B. iii., 154-164, 1905). 



I have added in italics the wave-length numbers corre- 

 sponding to Baly and Collie's oscillation frequencies for 

 comparison : — 



Vapour. Solution in alcohnl. , ^ ,,■ 



Friederichs. Hartley and Dobbie. Difference. Baly and Colhe. 



(1) 2670 



(2) 2633 



(3) 2588 



(4) 2526 



(5) 245S 



(6) 2404 



(7) 2356 



(8) 2305 



He points out that the bands of the substance in solu- 

 tion which without doubt correspond with those of the 

 vapour are all shifted towards the red, as might be 

 expected, but that the shift appears to be greater the 

 smaller the wave-lengths of the absorbed rays. The com- 

 parison of Baly and Collie's numbers with those of Hartley 

 and Dobbie is very interesting in this connection, inasmuch 

 as they show a close general agreement in their divergence 

 from the measurements of Friederichs. Furthermore, the 

 following points may be noted : — 



First, the omission of the second band in Hartley and 

 Dobbie's spectrum ; second, the omission of the eighth band 

 by Baly and Collie ; third, there is a close agreement 

 between Hartley and Dobbie's and Baly and Collie's 

 numbers in the first, fifth, sixth, and seventh bands, but 

 the two sets of measurements for the third and fourth 

 bands differ more widely than the others. 



It may be mentioned that the second very narrow 

 band is ' visible on the photographs taken by Hartley 

 and Dobbie, though it can scarcely be considered as 

 measurable; no doubt a longer exposure w-ould have 

 rendered it more plainly. Those who have measured 

 similar series of bands in the visible region, for example, 

 those in the spectrum of potassium permanganate, which 

 are also eight in number, will appreciate the close 

 approximation of the above figures. 



W. N. H.ARTLEV. 



Roval College of Science, Dublin, September 19. 



Rhymes on the Value of jr. 



The following rhyme is in imitation of the French 

 and German verses given in Nature (August 17) in which 

 the number of letters in each word correspond to a numeral 

 in the value of ir. The three concluding lines are some- 

 what obscure ; it seems to have occurred to the author 

 that the method is a misuse of language, and he expresses 

 the hope that N.ature will take a more lenient view than 

 Dr. Johnson might be imagined to express. 



To the Editor of N.^ture. 

 .Sir, — I send a rhyme excelling 



3 ' 4 I '5 9 



In sacred truth and rigid spelling. 

 26 5 3 5 S " 



Numerical sprites elucidate 



9 7 9 



For me the lexicon's dull weight. 

 323 8 46 



If " Nature " gain, 

 264 

 Not vou complain, 

 3 3 8 



Tho' Dr. Johnson fulminate. 

 327 9 



NO. 1875, "^'OI- 7-^ 



The Celtic Pony. 



In a review, signed " R. L.," of " The Frcroes and 

 Iceland," in Nature of September 21 (p. 506), I was 

 surprised to read that I had credited Prof. Ewart " with 

 being the first to regard Przewalslcy's horse as a variety 

 of Eqiius caballiis." I have just re-read the paragraph 

 relating to the wild horse in my " Appendix on the Celtic 

 Ponv," and I can find no passage which, it seems to me, 

 could by any possibility be made to bear this strange 

 construction. 



Sanson's subspecies E. c. hiberniciis appears to include 

 all the various ponies of the British Isles, the Breton in 

 France, as well as the horses of Iceland, Norway, and 

 Sweden. It has been recognised for some time past that 

 the Icelandic horses are of two different types, while the 

 Swedish horses are admittedly very mixed. Moreover, as 

 a result of a recent tour in Norway, it has become evident 

 to me that there are in that country at least two distinct 

 kinds of native horses (represented by the pure fjord horse 

 and the Gudbrandsdal horse). In view of these consider- 

 ations, the statement that the Celtic pony is " probably 

 inseparable " from the somewhat heterogeneous assemblage 

 (as it now appears to be) included under E. c. hiberniciis 

 becomes a little obscure. But, as " R. L." points out, 

 I did not make this statement. I grant, however, that it 

 might have been better had I made some allusion to this 

 matter. 



But why I should have been expected in an " Appendix 

 on the Celtic Pony " to have entered into a discussion as 

 to the proper technical name to apply to E. proewalskyi 

 or to have recorded an irrelevant criticism of Prof. Ridge- 

 way's new name of E. c. libycus, I am at a loss to under- 

 stand. Francis H. A. Marshall. 



The University, Edinburgh, September 24. 



GREEK ARCH.SOLOGY.' 

 T^HE archaeologist justly ranks himself as a con- 

 ■'- tributor to the worW's knowledge on the same 

 level as those who discover previously unknown forces 

 in nature or new facts in the life-history of animals, 

 extinct or living. Archaeology, which is a branch of 

 the great science of anthropology, discovers and cor- 

 relates new facts in the early history of civilisation. 

 Greek archaeological discovery must always be of 

 most especial interest, since it tells us of the origins of 

 that early civilisation of the Mediterranean basin 

 from which our present-day culture is derived. One 

 of the most welcome yearly publications dealing with 

 the subject is the " Annual of the British School at 

 Athens," the tenth volume of which lies before us. It 

 deals with the British work of 1903-4, besides con- 

 taining independent articles on matters of archre- 

 ological interest. 



Dr. Arthur Evans's work at Knossos does not 

 occupv so much space in the " Annual " as usual. 

 The discoveries of the year, while most interesting, 

 were not so new and epoch-making as those of former 

 years, and the chief find, the tombs of " Ja'far's 

 Papoijra " (toC T(a(f)ep fj Tlanovpa) and Isdpata, are 

 described by Dr. Evans in a separate communication 

 to Archaeologia. The first-named tombs, on a hill 

 north of the Knossian palace, were of various types; 

 (i) chamber-tombs approached by a dromos; "in 

 many cases these contained clay coffins, in which the 

 dead had been deposited in cists, their knees drawn 

 tow-ards the chin"; (2) shaft-graves; (3) pit-caves, 

 " or pits giving access to a walled cavity in the side 

 beloW'." In 2 and 3 the skeletons were extended 

 at full length. On the hill of Isopata, about two 

 miles north of Ja'f.^r's Papoijra, a very fine tomb, no 

 doubt that of a king, was found, with a smaller one 

 by its side. The larger consisted of a square chamber 

 of limestone blocks, eight metres bv six, " with the 



1 " The Annual of the British School at Athens," No. x. Session 1903-4. 

 (London : Macmillan and Co., Lid.) 



