542 



NATURE 



[Oct. 6, 18S7 



well-known experiments of Dumas it will be remembered that 

 an indefinite quantity of hydrogen was employed, the loss of 

 oxygen by the copper oxide and the weight of water formed 

 furnishing the only data obtainable by the then possible ex- 

 perimental methods. But Dr. Keiser has succeeded in weigh- 

 ing his hydrogen by taking advantage of its peculiar property of 

 being occluded by the metal palladium ; it is shown that a piece 

 of palladium ico grammes in weight will readily take up between 

 0"6 and 07 grammes of hydrogen, when heated in a stream 

 of the gas to about 150°. The palladium-hydroj;en compound 

 formed is perfectly stable at ordinary temperatures, and may be 

 preserved unchanged in a vessel filled with hydrogen gas ; but 

 on heating this qtiasi- sWoy the gas is slowly driven out again, 

 and by weighing before and after the heating, the weight of 

 hydrogen expelled may be accurately determined. This weighed 

 quantity of hydrogen gas was then passed over heated coppar 

 oxide, and the weight of water formed determined in the usual 

 manner. But Dr. Keiser has gone further than merely syn- 

 thesize water by the most direct means possible, he has refined 

 the process so far as to be enabled to employ it as a direct means 

 of determining the atomic weight of oxygen. The minutest 

 precautions were taken against error, and the purification of the 

 hydrogen carried out in a most thorough manner, with the un- 

 expected result that the atomic weight of oxygen is most prob- 

 ably slightly lower than I5"96 and more nearly IS'S/. It is 

 interesting to be reminded that Dumas states in his memoir: — 

 " Of all analyses, that of water involves the most uncertainty, 

 it is true that one part of hydrogen combines with eight parts 

 of oxygen to form water, and nothing could be nTore exact than 

 the analysis of water, if one were able to weigh the hydrogen 

 and the water formed by its combustion. Bat the experiment 

 is not possible in this form." However, owing to the ingenuity 

 of Dr. Keiser, this happy result has now been achieved. 



The Time; of September 22 contained an interesting account 

 of the Troglodyte remains in Southern Morocco. The diffi- 

 culties placed in the way of exploration in Morocco have pre- 

 vented an examination hitherto of these remains. The Troglodyte 

 caves are situated near Ain Tarsil, a village some three days' ride 

 to the south-west of the city of Morocco, near the borders of the 

 province of Haha. At the village the scenery undergoes a com- 

 plete change. Instead of the dreary plains one comes across 

 curious hills divided by great ravines, not unlike the canons of 

 California, in one of which is situated the strange city of the 

 Troglodytes. The gorge is a narrow one, the cliffs on either 

 side rising almost perpendicularly from the bottom of the deep 

 valley; after progressing some little way along this valley the 

 first caves are sighted. They are cut in the solid rock at a great 

 height from the ground, and are in some places single, in others 

 in tiers of two or three, one above the other. The entrances 

 are small, varying from 3 J to 4^ f<.et in height and about 3 feet 

 in breadth. In places where the rock has fallen away the 

 entrances are faced with masonry -of a neat and orderly type, 

 and in one or two cases where the natural formation of the rock 

 necessitated exceptionally large entrances this masonry served 

 also the purpose of dividing the door into two parts, one of which 

 no doubt served as a window. As to the means of access to 

 these caves, the writer discusses three theories, all possible, but 

 one alone of which is probable. The first is that the face of the 

 cliff has since the formation of the caves been worn away by 

 wind and water, and so crags and projecting pieces of rock that 

 once rendered access po5sible have disappeared ; secondly, the 

 Troglodytes may have been so active as to be able to get up 

 perpendicular smooth rock ; the third, that they used ropes and 

 ladders. He dismisses the first as unjustified by anything we 

 know of Morocco, and inconsistent with the existence of various 

 tiers of cliffs ; and as to the second, although it is stated that 

 these cave-dwellers were swift of foot as a hoTse, the shape of the 



cliff renders climbing an impossibility. The walls of rock are 

 perfectly smooth from the bottom to the height of the caves, 

 which is 200 to 300 feet in many cases, and it is only in a few 

 places that there are ledges sufficiently wide for birds of prey. 

 In favour of the ladder theory there is the circumstance that in 

 the doorways of many of the caves there still remains, a few 

 inches above the floor, and crossing from lintel to lintel, a bar of 

 wood some six inches in thickness, which must have acted as a 

 roller for ladders. The freshness of the wood is surprising, but 

 in the Karli Caves in the Western Ghauts there is a roof which 

 is considered by experts to be coeval with the caves themselves. 

 The writer was able to enter the caves in only one or two cases, 

 a landslip having built a pile of earth and stones that rendered 

 access possible. Passing through the low, narrow doorway, the 

 visitors found themselves in an oblong i-oom 15 feet by 7> lead- 

 ing from which at either end were smaller chambers. The wall 

 of the larger room adjacent to the cliff was pierced with three 

 windows ; each of the smaller ones also possessed a window. 

 The walls were rough, but the floor and ceiling were well 

 smoothed and cut. The height of each was about 5 feet 2 

 inches. In no cases were bones discovered, only a little broken 

 pottery and one or two doubtful flint-heads alone being found. 

 No doubt much greater success will await the explorer who 

 manages to overcome Moorish prejudice and bigitry, and by 

 means of ropes and ladders enters the upper tiers of caves which 

 have lain in their present state since the old race died out or 

 took to more rational abodes. The Troglodytes cannot have 

 been such savages as they are usually considered : their excava- 

 tions are hollowed out skilfully, and their abodes show signs of 

 great labour and some idea of care and comfort. 



The additions to the Zoological Society's Gardens during the 

 past week include two Macaque Monkeys {Macacus cynomolgus) 

 from India, presented by Miss Barker ; two Arctic Foxes {Canis 

 lagoptis) from the Faroe Islands, presented by Mr. T. Norden- 

 felt ; a Macaque Monkey {Macacus cynomolgus) from India, 

 presented by Mr. R. Taylor; a Corn-crake {Crex pratensis), 

 British, presented by Mr. Howard Bunn ; a Ring Dove 

 [Cohunba palumbus), British, deposited ; a Sumatran Wild 

 Dog {Canis javanicus'i ) from Sumatra, purchased; a Wedge- 

 tailed Eagle {Aquila audax) from Australia, received in ex- 

 change ; a Mule Deer {Cariacus tnacrotis), born in the Gardens. 



OUR ASTRONOMICAL COLUMN. 



Flamsteed's Stars "Observed but not ExisxtNG." — 

 Prof. C. H. F. Peters, in a paper published in vol. iii. of the 

 Memoirs of the National Academy of Sciences of Washington, has 

 discussed Flamsteed's observations of the twenty-t^vo stars which, 

 it will be remembered, Baily, in his " Account of the Rev. John 

 Flamsteed," entered in a list with the above heading, which, as 

 he explains, means stars "of which the observations appear to 

 be accurately recorded, but which still cannot now be found in 

 the heavens." As the disappearance of so many stars in com- 

 paratively so short an interval of time appeared to Prof. Peters 

 to be rather improbable, it seemed to him to be desirable that 

 these cases should be scrutinized with the additional means of 

 identification provided by the Durchmusterung, which of course 

 was not available at the time of Baily's publication. Acting, 

 then, on the assumption that some otherwise plausible error in 

 Flamsteed's entry of his observation which leads to a modern 

 star-place is to be held much more reasonable than the vague 

 acquiescence in a supposed disappearance of the star. Prof. 

 Peters considers that (making allowance for the probable error 

 of a position in Flamsteed's Catalogue) he has succeeded in 

 finding for every case at least a probable explanation. One of 

 the observations in question — that of the object No. 1647 in 

 Baily's " Flamsteed " — turns out to have been an observation of the 

 planet Uranus, an explanation which was suggested by Argelan- 

 der but rejected by Baily. The agreement, however, of the 

 position of the object observed by Flamsteed with the place of 

 Uranus computed from Newcomb's Tables is so close as to leave 

 no doubt of its identity. 



I 



