November 25, 1897] 



NA TURE 



Four tubes developed in bulb broad bands of light yellow- 

 green electrical molecules, interspaced with darker bands ; this 

 was like the marking of the planet Jupiter. 



Two tubes developed a mottled and leaf- patterned electric- 

 molecular condition, extremely like the appearance of the 

 photosphere of the sun (for the first time the skin of my hand 

 was affected by one of these tubes— November 1896 ; this hand 

 having some years previously had gouty eczema ; the other 

 hand was not affected in either case). 



Two tubes developed forms like the clouds known as " mares'- 

 tails " ; one tube, a form like the sulky lower strata of distant 

 thunder-clouds. 



The most interesting phenomenon was one which was also 

 seen by a medical friend : the whole tube was a mass of yellow- 

 green phosphorescence, even behind the kathode ; the molecular 

 film in glass bulb in front and round the edge of the kathode 

 (mottled condition) developed a small black spot the size of a 

 pin-head, which increased to the dimensions of a small pea, 

 broadened out into an irregular patch, split up into small spots, 

 which ran round the bulb, disappeared ; reformed into a patch- 

 diminished — and disappeared. The phenomenon was repeated 

 at regular intervals for ten minutes, then finally stopped. The 

 irregular patch was extremely like a sun-spot. 



From time to time I have called the attention of interested 

 friends to the above phenomena, so as to have witnesses, as the 

 tubes seldom repeat the same conditions, which can only be 

 attained by the application of heat while working the tube. 



The rays magnetised my watch on two occasions; an induction 

 coil having no magnetic field, owing to interrupted current, 

 could not do this. William Webster. 



Art Club, Blackheath. 



REMARKABLE TERMITE MOUNDS OF 

 AUSTRALIA. 



'T^HE destructive propensities and architectural en- 

 ■*■ dowments of the termites or white ants are 

 familiar subjects to most travellers and residents in 

 tropical countries. Notwithstanding, however, the almost 

 cosmopolitan distribution within tropic areas of these 

 insect pests, an astonishingly small amount of accurate 

 data has been chronicled concerning their specific varieties 

 or the widely varying modifications of their social tene- 

 ments. 



As a matter of fact the figures and descriptions of 

 the insects and nest mounds or "termitaria" of the 

 African white ants Termes bellicosus, T. mordax, and 

 other species contributed by Henry Smeathman to the 

 Transactions of the Royal Society for the year 1781, con- 

 stitute up to the present date the standard account of 

 white ants and their ways that is reproduced with trivial 

 variations in most modern zoological text-books. Much, 

 undoubtedly, has been accomplished within recent years, 

 notably through the investigations of Grassi and Sandias, 

 Fitz Miiller, and other biologists to elucidate the minute 

 anatomy, individual modifications, and social economy 

 and relationships of the indigenous or imported South 

 European types Termes incifugus and T. flavipes. 

 Neither of these forms, however, are mound constructors, 

 but live within subterranean tunnels or in excavations of 

 the wood which they devour. Smeathman is hence still 

 left, facile princeps, the first, almost the only authority 

 on the architectural fabrications of the mound-construct- 

 ing species. 



While Africa in connection with Smeathman's investiga- 

 tions has hitherto inherited an undisputed monopoly in the 

 matter of termite mounds, it will probably be found that the 

 island continent of Australia can produce an equally, and 

 in some respects even more noteworthy variety of these 

 insect edifices. So far as altitude is concerned, the palm 

 among the Australian series must be conceded to the 

 huge buttressed columnar termitaria that occur some 

 forty miles inland from Port Darwin in the Northern 

 Territory of South Australia. A height of at least eighteen 

 feet is not unfrequently attained by this variety, and their 

 diameter being almost equal throughout, their appearance 



NO. 1465, VOL. 57] 



has been appropriately likened to miniature towers. 

 Termitaria almost equal in altitude to this Port Darwin 

 variety have also been observed by the writer in York 

 Peninsula, North Queensland, and are notably abundant 

 on the pomt of land abutting upon the mainland fore- 

 shore of the southern entrance to the Albany Pass. The 

 largest termitaria of this type occur, however, in the 

 vicinity of the bridle path between Mr. Frank Jardines 

 homestead at Somerset, and the former telegraph station 

 at Patterson facing Endeavour Strait. Passing them on 

 horseback, many of these mounds were observed to 

 tower considerably above the rider's head. The contour 

 of these York Peninsula termitaria differs essentially 

 from that of the Port Darwin form. In place of being 

 subcylindrical or columnar, they are distinctly pyramidal, 

 widest at the base, and tapering up to a single, or it may 

 be occasionally several accuminate points. 



Within the tropical area of Western Australia, known 

 as the Kimberley district of that colony, a third large 

 and very distinct type of termitarium is dominant. 

 While most commonly presenting a symmetrically 

 conical or hemispherical contour, it not unfrequently 

 happens that these white ant mounds are built up into 

 the most irregular and fantastic shapes. Throughout all 

 such innumerable modifications, however, one essentially- 

 unique fundamental plan of structure is predominant. 

 This is exemplified by the circumstance that each nest 

 mound or termitarium presents the appearance of having 

 been built up, as it were, by the superposition of con- 

 secutive hod-loads of half solidified mortar, and which 

 before setting has partly overflown and overlapped the 

 preceding instalments. 



This lobulated or Kimberley type of termite mound, 

 as it may be designated, occurs in great abundance and 

 in its finest development in the thinly wooded scrub or 

 so-called " pindan " in the neighbourhood of Derby, at 

 the head of King's Sound. One of the tallest of these 

 termitaria vyas found to measure as much as fourteen feet, 

 but this altitude is occasionally exceeded. The separate 

 occasions on which the writer visited King's Sound were 

 utilised by him for obtaining information concerning the 

 approximate time that is occupied by the termites in the 

 reconstruction of a partially dismantled edifice. Examples 

 of termitaria about eight feet high were bisected vertically, 

 the one half being completely demolished. Within 

 twelve months one quarter of the denuded area had been 

 filled in. Visited eighteen months later, fully two-thirds 

 of the demolished moiety had been rebuilt, and it would 

 evidently have not taken more than another twelve- 

 months, or a total of between three or four years, to 

 restore the mound to its original symmetrical shape and 

 dimensions. 



The most remarkable, though by no means the most 

 lofty of the termite edifices peculiar to Australia, are 

 those modifications of these structures which are popu- 

 larly known by the local titles of "meridian," "magnetic," 

 or "compass" ants' nests. The termitaria belonging to 

 this category are distinguished by their elongate, much 

 compressed contour, comparable in the most evenly con- 

 structed examples to huge slabs of undressed sandstone 

 set up on edge. The upper edge or ridge of the 

 "Meridian" termitarium is always the narrower, and is 

 either nearly smooth, irregularly serrated, or may be de- 

 veloped into a series of slender pinnacles. The most 

 notable feature respecting these white ant mounds is, 

 however, the circumstance that the orientation of their 

 longer axis is always coincident with the local parallel 

 of latitude, or, in other words, in a precise line with the 

 north and south points of the magnetic compass. 



The most striking examples of these meridian ants' 

 nests observed by the writer occur in the Laura Valley, 

 North Queensland, some sixty miles from Cooktown. 

 The altitude of these Laura Valley ant mounds is not 

 very considerable, rarely, indeed, exceeding six or eight 



