RevieiDS — Bernard's MadreiJorarian Corals. 465 



by Mr. Brook under 180 specific names, the total number of species 

 of the genus amounting to 221. Vol. ii, embracing the genera 

 Tarhinaria and Astrceopora, was, through the lamented death of that 

 accomplished zoologist Mr, George Brook, undertaken by another 

 excellent biologist, Mr. Henry M. Bernard, M.A., F.L.S., F.2.S., 

 and appeared in 1896 (pp. iv and 106, with 33 plates). Vol. iii, 

 containing the corals of the genera Moatipora and Anacropora, 

 prepared by the same author, was issued in 1897 (pp. viii and 192, 

 with 34 plates 4to). In 1897 the late Sir William Flower, then 

 Director, writing in the Preface, said : " The three volumes now 

 finished form together a very complete monograph of the Madre- 

 poridse, which is one of the chief reef-building families of Stony 

 Corals. . . . In 1878, when Briiggemann prepared his manuscript 

 catalogue, there were only 41 specimens of Montipores, divisible 

 into 16 species. There are now more than 400 specimens, classed 

 under 116 species, about 80 of which are new." 



In the preface to vol. ii Mr. Bei-nard refers (pp. 19-21) to 

 the difficulties which beset the systematic naturalist in attempting 

 to establish a standard upon which to divide Corals into groups — 

 say according to their methods of growth. These methods of 

 growth, which at first seem so convenient, are found to pass into 

 one another, so that it is most perplexing to decide whether 

 a specimen exhibits one or the other type of growth. 



Again, the number of transitional forms observable in a long 

 series of specimens, which, from their calices, are clearly related, 

 renders it difficult to decide as to which group they properly belong. 

 Still more serious is the fact that in Torres Straits we find 

 Turbinarians widely difi'ering in the character of their calices, yet 

 revealing exactly the same methods of growth, which shows that 

 in that case at least the form of the corallum is due to the 

 environment. Often, too, special local forms of growth are con- 

 fined entirely to limited areas. 



A year later, writing at p. 17 in vol. iii, Mr. Bernard says: 

 " While claiming that the chief divisions of the genus are natural 

 divisions, I can only repeat what was said in the preface to vol. ii 

 as to the real value of the specific divisions. The types represent 

 merely the more marked variations presented by the specimens in 

 the collection, and are therefore for the most part purely artificial 

 groupings. Only in those cases in which the individual specimens 

 are known to have been collected from the same locality, and might 

 almost be fragments of one and the same colony, does the name 

 imply the close blood-relationship which the word species should 

 be taken to connote. In all other cases the types are, strictly 

 speaking, only morphological groups united because of certain 

 peculiarities of form or structure which they have in common. 

 Their ultimate systematic value is thus problematical. How much 

 this is the case, indeed, may be gathered from the fact that the 

 differences presented by specimens which are undoubtedly specifically 

 identical may be far more striking than those that separate many 

 of the types," 



DECADE IV. — VOL. X. — XO. X. 30 



