Nov. 25, I 



NA TURE 



77 



further from the pheasant, so that these would be quite infertile. 

 But this is merely argument from analogy ; there is no evidence 

 of the result of such superposed "sports," and retrogression to 

 greater fertility seems possible. 



This instance is not a good one, because (he observed partial 

 infertility (i.e. only occasional fertility) between pheasant and 

 b'ack grouse may be due to dislike, not partial impotence. But 

 I doubt whether distaste for pairing and impotence when paired 

 are often quite dissociated. 



Mr. Murphy asks, as I asked in these pages in 1884, and others 

 have asked since, for one simple fact which would be decisive. 

 Is it, or is it not, the fact that allied species which are confined 

 each to a particular island, prove, when brought together, far 

 less frequently infertile than species, equally dissimilar, which 

 had lived in the same district, might be expected to prove. On 

 the answer to this question depends, as far as I can see, the fate 

 of the theory of physiological selection. Can no one answer it ? 

 Ed.mund Catchpool 



THE CORAL REEFS OF THE SOLOMON 

 ISLANDS 1 

 /^UT of a collection of nearly seventy corals which I 

 ^-^ made in these islands, nearly a quarter are new or 

 undescribed ; and from this fact, as I am informed by 

 Mr. S. Ridley, it may be inferred that there is yet much 

 to be learned of the corals of this region. After describ- 

 ing in my paper the characters of a typical reef, I pro- 

 ceeded to refer to the complex relations that exist between 

 the multitudes of creatures that frequent coral reefs. The 

 protective colouring of the small crabs that live among 

 the branching corals often attracted my attention. I re- 

 call, in particular, the instance of a small crab that finds 

 its home among the branches of a Pocillopora. The light 

 purple colour of its carapace corresponds with the hue of 

 the coral at the base of the branches, where it lives ; 

 whilst the light red colour of the big claws, as they are 

 held up in their usual attitude, similarly imitates the 

 colour of the branches. To make the guise more com- 

 plete, both carapace and claws possess rude he.\agonal 

 markings, which correspond exactly in size and appear- 

 ance with the polyp-cells of the coral. Another species 

 of crab, that climbs about the blue-tipped branches of a 

 Madrcpora, has the points of its pincer-claws similarly 

 coloured. It is interesting to note that these two crabs 

 are adapted to live each on its own species of coral. Had 

 I caused them to exchange their homes, their borrowed 

 hues and markings would have at once made them con- 

 spicuous objects for their enemies. 



I paid especial attention to the inter-tidal exposure of 

 living corals, and was much surprised at the number of 

 species which are bared by the ebbing tide. Of all 

 the corals in these islands, those belonging to the genus 

 Ca'loria seem to be the hardiest in this respect. In the 

 paper I have described my observations with some detail. 



Coral Reefs and Shoals. — The earliest condition of the 

 coral reefs in this group is to be found in that of the 

 numerous detached submerged reefs or shoals which lie 

 below the limit of the constructive power of the breakers, 

 having been arrested in their upward growth at depths 

 varying between 5 and 10 fathoms according to the 

 exposed or protected character of their situation. This 

 remarkable fact of the arrest of the upward growth of the 

 coral at these depths was utilised by Lieut. -Commander 

 Oldham whilst surveying these svibmerged reefs in 

 H.M.S. Lark. If a shoal was not marked at the surface 

 by a reef-flat or by an islet, we could sail over it with 

 perfect safety. The broken water of the tide-rip that 

 indicated these shoals was no source of danger for a 

 vessel of light draught. In my paper I have given evi- 

 dence to prove that a shoal which was found by Bougain- 

 ville in 1768 to be covered by 5 fathoms, remains in the 



' Abstract of aPaperby H. B. Guppy, M.B., late Surgeon H.M.S. Latk. 

 Communicated by Dr. John Murray to the Royal Society of Edinburgh on 

 July 5, 1886. 



same condition at the present day. The nutnber of coral 

 shoals possessing these characters led me to the con- 

 clusion that isolated submerged reefs are unable without 

 the assistance of a movement of elevation to raise them- 

 selves to within the constructive power of the breakers. 

 When they have reached their upward limit, they extend 

 laterally, forming ultimately flat-topped shoals. It may 

 appear bold to suggest that atolls and barrier-reefs owe 

 their appearance at the surface to a movement of eleva- 

 tion ; but we know that in the regions occupied by the 

 atolls of the Low Archipelago, of the Fiji Islands, and of 

 the Pclew Group, the last movement experienced has 

 been one of elevation ; whilst the observations of IVIr. 

 Beete Jukes on the Australian Barrier-Reef go to show 

 that, if there has been a recent change of level in that 

 region, it was one of the same nature. In the atoll of 

 Oima in the Solomon Group I found evidence of an 

 anterior elevation. 



In my paper I proceeded to describe at some length 

 the reefs that have reached the surface. In this abstract, 

 however, I can only refer to the fact that all the three 

 classes of reefs are to be found in this group ; the atolls, 

 I should add, being comparatively few in number and of 

 small size. 



The Formation of Atolls. — My observations go to 

 show that atolls of small size (a mile or two across) do 

 not assume their characteristic form until they have 

 reached the surface. After upheaval has brought a sub- 

 merged coral shoal within the constructive power of the 

 breakers, it soon appears at the surface as an isolated 

 patch of reef. Extensions or wings grow out on either 

 side, and, guided by the prevailingcurrents (in the manner 

 described by Semper), they ultimately form the common 

 horse-shoe reef, which presents its convexity against the 

 currents. Large atolls evidently begin to assume their 

 characteristic shape below the surface as described by 

 Murray and A. Agassiz. 



Tlie Formation of Barrier-Reefs. — The facts on which 

 my conclusions have been based were obtained by the 

 examination of the weather slopes of reefs. For the 

 first 70 or 80 yards from the weather edge of a reef 

 there is a gradual slope, largely bare of living coral, to a 

 depth of 4 or 5 fathoms. There is then a rapid descent 

 to a depth varying between 12 and 1 8 fathoms. It is this 

 declivity that constitutes the growing edge of the reef, 

 and the sand and gravel produced by the constant action 

 of the breakers collect at its foot. When the submarine 

 slope is more than 10° or 12°, as is usually the case, the 

 sand and gravel extend far beyond the depths in which 

 reef-corals thrive ; but when the slope is gradual, i.e. less 

 than 5°, the lower margin of this band of detritus lies 

 within the reef-coral zone, and in consequence a line of 

 barrier-reef is ultimately formed beyond this band with a 

 deep-water channel inside {vide diagram). Should the 



Barrier-reef of Choiseul Bay (drawn on a true scale to the loo-fathom line. 

 a = incipient barner-reef (size purposely exaggerated) : b = belt of sand 

 and guvel. 



area be undergoing elevation, a succession of concentric 

 lines of barrier-reefs will originate, line after line being 

 advanced as fresh portions of the sea bottom are brought 

 towards the surface, each line growing upward along the 

 lower margin of the band of detritus derived from the line 

 of reef inside it. In such a manner have the Shortland 

 Islands been produced. When I arrived at the above con- 

 clusion I was riot aware that substantially the same 

 explanation had been advanced thirty years before by 



