TROPICAL BAPTISM 23 



Like all not very recent madreporic formations the islands 

 and coasts of the Massawa-Dahlak region are 'belted' con- 

 sistently by a sandy platform (or terrace) strewn with 

 madreporic blocks of no great size (about two or three yards 

 maximum in diameter) . The edge of this terrace can thrust 

 itself into the sea for some dozens or even hundreds of yards, 

 and is almost always marked by a *crown', or barrier of 

 living coral. The barrier falls fairly steeply, sometimes per- 

 pendicularly, for twenty to fifty feet on to a new bed of 

 sand, the ulterior slope of which descends smoothly and not 

 too rapidly towards the succeeding and decisive shelf (or 

 drop) that ends the continental plateau, and from there 

 sinks hundreds of yards into the abyss. 



Our studies and research were to be pursued in the littoral 

 sector of the platform and the madreporic barrier (a sector 

 comprising a depth of about sixty feet). In this zone, and 

 above all among the corals, the equipment of non-underwater 

 naturalists has never brought any great success ; nets get torn 

 to pieces on the branches of the petrified forest, and fish- 

 hooks catch relatively few fish of relatively few species. We, 

 on the other hand, could use a very difierent technique, and 

 our results did in fact surpass all hope or expectation. 



There I was then, face to face with the famous coral 

 barrier for the first time. I must confess that it looked a 

 pretty miserable thing to me and I was almost disappointed. 

 The more twisted and ramified corals formed, it is true, an 

 entanglement astonishing in its morphological variety, but 

 when I decided to go down as far as the base of the shelf to 

 search for the sand, I knocked my nose on the bed of the 

 sea about twenty feet down. The barrier was only fifteen to 

 seventeen feet in height, a thorny brown wall in the turbid 

 water. 



But it was a wall that moved, that swarmed. Again I went 



