1909.] N. AnnandaIvE : The Indian Cirripedia Pedunculata. 67 



is artificial. The question need not arise as regards genera containing a small number 

 of species, for in such cases it is clear that the multiplication of divisions is unnecessary 

 and apt to confuse rather than elucidate. In the case of species, however, I am in 

 favour of only recognizing those which are clearly different from one another and 

 separated by a distinct gap in the line of variation. Narrow gaps as regards structure, 

 which cannot be bridged over completely, exist between the offspring of a single parent 

 or pair of parents, as we see in the larger animals, in which they are easily detected; 

 and narrow gaps, a little wider in many cases, are clearly visible between the children 

 of different parents. It is impossible to take notice of all such gaps, many of them 

 being incapable of detection by the senses and rather inferred than proved to exist, 

 while others, which can be detected, are so inconstant, and obey any known law so 

 little, that to recognize them causes confusion and does not assist in a rational 

 system of classification. There are other narrow gaps which can not only be detected 

 but also proved to separate large groups of individuals to some slight extent. They 

 are still so narrow, however, that they can only be detected by the closest study 

 (or even by a comparison of specimens), and are not sufficiently marked to be used as 

 arguments in proving any divergence in genetic relationship between the groups they 

 separate. 



Groups thus separated are peculiarly common in the Cirripedes and occur in many 

 species of Lepadidse. To regard them as distinct species would militate against any 

 investigation of the geographical distribution of the group, and would only complicate 

 its study from a morphological point of view. It must be clearly understood, more- 

 over, that the gap which separates them is not of the same nature in all cases, or due 

 to the same causes ; and therefore it is inadvisable to call all of them either ' ' varie- 

 ties," '* subspecies," or '' races," or to designate them in any other way that fails 

 to mark their divergence in character. 



Apart from developmental phases, which have been little studied in the Lepadidae 

 and are not always easy to distinguish, we find that the smaller groups into which 

 the species may be divided are mainly of two kinds. In the first place it is easy to 

 prove that certain species are peculiarly liable to give rise to local races, the indivi- 

 duals of which are more or less constant among themselves and differ from individuals 

 from any other locality. Such groups I have called races or subspecies. Another 

 kind of group, however, occurs not infrequently, consisting of a number of individuals 

 not separated by any geographical boundary from others of the same species but 

 differing from the typical form {i.e., the form first described) in unimportant charac- 

 ters, which may or may not be hereditary. Such groups I have called '' varieties." 

 In most cases there is nothing definite to prove whether the varieties of a species 

 do or do not interbreed, but the probabilities are in favour of the former mode of 

 propagation. 



The best example of a species with different races that I know among the 

 Lepadidae is Pcecilasma kcempferi, which occurs in its typical form in Japan and the S. 

 Pacific but is represented by subspecies in several different parts of the Indian and 

 Atlantic Oceans, True varieties are less common among the Lepadidae than they are 



