REVIEW OF HIPPOCAMPUS — GINSBURG 507 



Besides the difference in the number of angles, the extra plates to a 

 pronounced extent also cause another effect, namely, the relation 

 between the upper ridge of the tail and that of the trunk. The extra 

 plates occupy a position above the upper plates of the tail segments. 

 The longitudinal ridge formed by the extra plates is, however, con- 

 tinuous or nearly so with the upper ridge of the trunk. This effect is 

 caused by one or two of the upper plates of the trunk being pushed 

 upward to form a continuous ridge with the extra plates. For in- 

 stance, if the extra plates are present on the first caudal and last 

 trunk segments — this being the usual condition in the subgenus 

 Hippocampus — the penultimate trunk segment has the position of the 

 upper plate raised to form a nearly continuous ridge with that of the 

 extra plates. On account of the presence of extra plates, the upper 

 ridge of the trunks is raised sharply above and overlaps that of the 

 tail. The tail ridge is continuous, or very nearly so, with that ridge 

 on the last one or two trunk segments, w^iich is immediately below 

 the ridge formed by the extra plates. Consequently, the number of 

 segments on which the two ridges overlap corresponds to the number 

 of extra plates under the base of the dorsal. 



It is now evident that the modification of the structure may be 

 described in three different ways: (1) By the number of angles in each 

 of the modified segments; (2) by the number of segments on which the 

 two upper ridges overlap and the position of those segments; and (3) 

 by the number of extra plates and their position. No matter what 

 form of descriptive statement is chosen, it is to be remembered that 

 it refers to the same fundamental structure, namely, the presence of 

 extra plates, their number, and position. In the diagnoses that follow 

 under each species, the variation is described first with reference to 

 the number of angles in each modified segment, and statements are 

 added in regard to the extra plates and the overlap of the upper ridges 

 of the tail and trunk. All three forms of the statement, however, 

 refer to the same fundamental structure, which is of considerable 

 importance in classification. 



Sometimes, as an individual variation, a segment may have an 

 extra plate on top on one side only. For instance, in a species in which 

 the first caudal segment is normally hexangular, an individual fish 

 may lack the extra plate on one side only. Such a segment is here- 

 after designated as incompletely hexangular, octangular, or novem- 

 angular, as the case may be. While such a designation, literally 

 speaking, is incorrect, it yet describes the fundamental condition and 

 variation of the structure. 



Proportional measurements. — The adoption of a uniform system in 

 taking some of the measurements is especially important in the sea- 

 horses, since these fishes assume an infinite variety of forms, on 

 account of the flexibility of the tail and the "throat" region. In 



