2 MADREPORARIA, 



differentiated. In this case, it was found that owing to the morphological obscurity of the 

 family, over twenty-seven genera, recent and fossil, had been from time to time incorporated 

 into it. Many of these gave a great deal of trouble. An illustration will make this clear, and 

 at the same time exemplify the method by which the difficulties were overcome. 



The genus Alveopora has a fragile, porous skeleton, which gives it a superficial claim to 

 be a Poritid. This claim, founded originally on the famous trabecular theory of coral morpho- 

 logy, which formed the basis of the great work of Milne-Edwards and Haime, becomes strong 

 when it is found that specimens of the chief Poritid genus Goniopora simulate the skeleton of 

 Alveopora so closely * that it was possible for anyone to maintain that the two genera passed 

 into one another. Those who doubted the relationship had apparently no more ground for 

 doubting than those who af&rmed it had for affirming it. It was a mere question of similarity 

 between stray specimens on one side, and dissimilarity of stray specimens on the other. This 

 question as to whether the Poritidae could or could not include the genus Alveopora had, then, 

 to be decided one way or the other before any clear conception of the essential morphology of 

 the family could possibly be arrived at. Preliminary draft catalogues of all the specimens in 

 the National Collection, first of the genus Goniopora, then of Alveopora, and lastly of Porites, 

 had to be prepared. Only in the course of this work did the distinguishing morphological 

 details stand out clearly enough to enable this volume to be written. 



In passing, it may be stated that the analysis of the genus Alveopora resulted in the 

 confirmation of the view which regards it as a survival of the Palseozoic Eavositidse, a family 

 wliich, except for the genus Alveopora, has long become extinct. f The place, then, of this genus 

 in this catalogue must be in the series dealing with the most primitive forms and nowhere 

 near the Poritidae, which stand at the very head of the tribe as the most recent and specialised. 



The results of the work on Goniopora and Porites will be found in this and in the next 

 volume (Vols. IV. and V.). I may add that the final separation oi Alveopora from the Poritidaj 

 was by no means the only question which had to be solved. For instance, Goniopora and 

 Porites were clearly both Poritids. But the relationship between them has only become clear 

 to me during the last year, after I had reluctantly begun to prej^are these pages for press, 

 having decided to leave this relationship an unsolved problem. For the solution, see Sec. IV. 

 p. 27. 



Then again, the difficulty of unravelling the morphology of the genus itself may be 

 gathered from the fact that an apparently perfectly simple question has puzzled every student 

 of the genus in turn (see Historical Sketch). The problem is : which are the older 

 caUcles, those at the top or those at the side of a massive stock ? The solution here arrived 

 at (see Sec. III.) was most unexpected, and there is no wonder that the problem has been a 

 standing difficulty. It can hardly fail to have a fascination for every student of coral 

 morphology. 



In addition to these complex morphological problems, the solution of which has so delayed 

 the appearance of this volume, a difficulty of a totally different character arose and had to be 



* Cf. e.g. G. Great Barrier Reef 10, p. 56. t See Journ. Linn. Soc, xxvi. 1898, p. 495. 



