550 



NATURE 



[April s, 1888 



much of the loose soil of a coral islet can be moved by a single 

 tidal encroachment. It happened that I was riding past the 

 very thin strip of land between Minni Minny and Barton Point 

 the day after an abnormally high tide. The strip of land here 

 is not more than 30 yards across, and the sea had washed right 

 over it on the previous day, clearing away an amount of soil 

 which was almost incredible. My companion, M. Casimir 

 Leconte, told me that the sea had not been known to wash o.ver 

 this place before. It was apparent that, after a few more of such 

 high tides as I had witnessed, a permanent breach would be 

 made at this spot, and another lagoon outlet would be formed, 

 which would be continually deepened as the tide set through it. 

 At the south-eastern side of the island I noticed that the land 

 was being rapidly destroyed on the outer shores just opposite to 

 a half-formed barachois, whose margins are situated not 60 

 yards from the outer shore. If the same process of external 

 destruction continues, whilst the barachois is deepened and 

 scooped out from within, it will not be many years before the 

 ocean makes a new channel into the lagoon at this point. Thus 

 the continuous strip of land which now nearly encircles the 

 lagoon of Diego Garcia is tending to be split up again into a 

 series of islets. At the points where the breaches are made the 

 tides and ocean currents will rush with great force into the lagoon, 

 and will scour out deep channels similar to that now existing 

 between Middle and East Islets. 



These facts taken together show how the normal action of 

 tides, winds, and waves is constantly tending to lower to the sea- 

 level any dry land that may have been formed by elevation or 

 otherwise. It does not seem to me to be surprising that the 

 majority of atolls and barrier reefs are, under such circumstances, 

 only just able to maintain their surfaces above the sea-level. 



No explanation of atoll formation would be complete if it did 

 not include an explanation of the Maldive atolls. This has been 

 felt by Darwin, who has explained the formation according to 

 his theory. Without attempting to enter into a lengthy discus- 

 sion, I will give my own explanation of the atoll. Tilla-dou- 

 Matte atoll is, as is well known, a huge atoll composed 

 of atolls. The islets forming the rim of the main atoll are 

 themselves atolls with their own lagoons ; the main lagoon 

 contains a few secondary atolls corresponding to the coral 

 patches in an ordinary atoll. It M'ill be generally admitted 

 that coral reefs are constantly increasing to seaward because 

 of the excessive growth of coral on their external slopes. ^ 

 As the inward shores of an atoll are constantly being 

 removed, and an atoll if completely formed tends to be broken 

 up again into small islets when it has reached a certain size, and 

 as the channels between the islets must be continually deepened 

 by the scour of the tides until deep passages are formed, an atoll 

 like Diego Garcia may be expected to reach in time a condition 

 like that of Peros Banhos. It is probable that a large bank like 

 the Great Chagos Bank, when it reaches the surface, can liever 

 give rise to a continuous strip of land, but must consist of a chain 

 of islets separated by channels of some depth and by tracts of 

 submerged reefs. The islets and tracts of reef in either case 

 would be bounded by deeper channels, and these channels, swept 

 by strong currents, would become wider and deeper, for corals 

 could not thrive in them. After a time the islets would become 

 so far isolated, and the entries into the lagoon would become so 

 large and numerous, that oceanic conditions would prevail in the 

 lagoon, and then there would be around each separate islet or 

 piece of reef all the necessary conditions for the formation of a 

 new atoll. The currents would impinge upon one side of the 

 islet or reef, sweep round it, and give a backwash at the further 

 side ; the corals would flourish in the circumferential parts of the 

 reef surrounding the islet, and new atolls with shallow lagoons 

 would be formed. 



In Tilla-dou-Matte the lagoons of the secondary atolls are 

 tolerably deep. In this case they must have been formed before 

 any land reached the surface. Applying the same reasoning as 

 in the former case, it can readily be understood how in the case 

 of the Great Chagos Bank, which has wide and deep breaches 

 in many places, the isolated reefs as they grow to the surface 

 must tend to assume an atoll form. An examination of the chart 

 shows that this is the case. The Great Chagos Bank in the course 



' This statement may at first sight seem at variance with what I have just 

 said about the rapid destruction of land on the outer and inner shores of an 

 atoll ; but in the latter case it is land above ivater that is destroyed. Coin- 

 cidently with this process the reef rock below water is constantly tending 

 to raise itself and to spread in all directions, owing to the perpetual growth 

 of corals and the accumulation of their skeletons. 



of time will rise to the surface as an atoll composed of secondary 

 atolls or atollons, similar to, but on a smaller scale than, the 

 Tilla-dou-Matte atoll. The explanation of atollons in the centre 

 of a large lagoon in which oceanic conditions have been 

 established, is quite obvious. 



THE ROYAL HORTICULTURAL SOCIETY. 



C\^ Tuesday, March 27, the Scientific Committee of the Royal 

 ^-^ Plorticultural Society met in the Committee-room of the 

 Drill Hall. Among the numerous subjects brought forward 

 were the followin;; : — 



Dispersal of the Seed in Pinus insignis. — Dr. ^Masters, 

 alluding to the great differences that exist in the species of Pinus, 

 as to the time at which the constituent scales of the cone 

 separate in order to liberate the seed, showed a series of cones 

 of Pinus insignis, the oldest of which bore the date 1S64. 

 In this all the scales were widely separate. The most recent 

 cones dated from 1877, and in them the scales were not at 

 all separated. Between these two extremes, cones were shown 

 exhibiting almost every intermediate stage of separation. It 

 is to be remarked that the separation begins generally just 

 above the centre of the pendulous cone on the side furthest 

 away from the branch, at the place where the excentricity of the 

 cone, due to the free exposure to light and air, and the absence 

 of obstacles afforded by the branch was greatest, and that it 

 follows a spiral course towards the base of the cone. The scales 

 separate in successive spiral coils, till, at length, all except a few 

 at the base and apex respectively, and which are probably sterile, 

 are separated one from the other. 



Semi-double and other Orchids. — Dr. Masters explained 

 the construction of numerous malformed orchids which were 

 interesting as throwing light on the morphology of the order. 

 Some extraordinary malformations of Fuchsias were shown, 

 and a drawing was exhibited of a magnificent new Anthurium, 

 which had appeared accidentally with an importation of Cattleya 

 Gaskelliana, in the garden of the Right Hon. J. Chamberlain. 

 The heart-shaped leaves are of gigantic size, and the large boat- 

 shaped spathe is of the richest crimson colour. 



Eucalyptus urnigtra. — Dr. Masters showed specimens of this 

 Tasmanian species in flower and fruit. They had been received 

 from Whittinghame Gardens, Prestonkirk, near Edinburgh, not 

 far from the sea, and where the tree is perfectly hardy. 



Daffodil with Crested Corona. — Rev. E. C. Gabbett sent 

 through Dr. Masters two flowers of a curious Daffodil from plants 

 growing on his lawn in Co. Limerick. The "frill," or outgrowth, 

 is produced from the outer surface of the corona, which has 

 thus a very peculiar appearance. 



Douglasia Icevigata. — Mr. G. F. Wilson alluded to this plant 

 as having been shown for the first time. It is a low-growing 

 Primulaceous plant, with tufted leaves and lilac flowers, like those 

 of an Androsace, but larger, and with the tube of the corolla 

 longer than the calyx, and with only two seeds to the capsule. 

 The species are the natives of North- Western America, the first 

 known species having been collected by Douglas not far from the 

 sources of the Columbia River, and named in his honour by 

 Dr. Lindley. 



Araucaria imbricafa Timber. — Mr. Ford, gr., Leonardslee, 

 exhibited slabs of wood cut from a tree of this species, and 

 which at 6 feet from the ground girthed 26 inches, the tree being 

 35 feet in height. The wood was yellow, soft, evenly grained, 

 and, judging by the distance between the rings, quickly grown. 



Numerous other plants and objects of interest were exhibited 

 and commented on. 



SOCIETIES AND ACADEMIES. 

 London. 



Royal Society, March 15. — "A Class of Functional 

 Invariants." By Mr. A. R. Forsyth, F.R.S. 



The memoir is occupied with the investigation of a class of 

 functional invariants, constituted by combinations of the partial 

 differential coefficients of a dependent variable, z, with regard 

 to two independent variables, x and y. The definition of the 

 invariant is given by the property that, when the independent 

 variables are transformed to X aad Y, and the same combina- 



