Mabch 29, 1907] 



SCIENCE 



50^ 



number of sexually produced generations 

 which are possible every year. Polyarthra 

 platyptera, a common cosmopolitan rotifer, 

 produces, in the Illinois River, male and 

 ■winter eggs throughout most of the year at 

 the close of parthenogenetic cycles of 3-6 

 weeks duration. It is not improbable that 

 other organisms of the same habitat run a 

 similar course and exhibit a corresponding 

 proportion of sexual reproductions. Oppor- 

 tunities for swamping out incipient species 

 are present, therefore, in these organisms, at 

 least in so far as the existence of sexual 

 reproduction oflFers them, in limnetic and 

 pelagic organisms as well as in the fauna and 

 flora of the land or in animals of higher 

 organization. The work of Maupas and 

 Calkins on the recurrence of conjugation in 

 ciliates is significant as indicating the neces- 

 Bity of sexual reproduction among unicellular 

 organisms. 



Evidence of the coincident distribution of 

 related species of higher pelagic organisms in 

 which amphigony is the only form of repro- 

 duction may be found among the chstognaths. 

 Other groups of pelagic organisms should be 

 examined on this point. The publication of 

 Fowler's excellent monograph on the chsetog- 

 naths in the reports of the " Siboga " expedi- 

 tion, in which he reviews the matter of their 

 distribution, affords some pertinent data of 

 prime significance because the group is a 

 sharply circumscribed one with but few 

 species, the distribution of most of which is 

 fairly well known. There are but three 

 genera, Sagitta, Krohnia and Spadella in 

 which the author recognizes twenty, three and 

 two species, respectively. The two species of 

 Spadella are members of the epiplankton and 

 their areas of distribution overlap. In the 

 case of Sagitta there are twelve species which 

 belong to the epiplankton of warm temperate 

 and tropical waters. In this group are both 

 cosmopolitan and apparently restricted spe- 

 cies. The remaining eight species belong to 

 colder or deeper waters (5) or are of unde- 

 termined range (3). Of the twelve species at 

 least ten are known to occur in the Indo- 

 Australian or Indo-Pacific regions, where they 

 exhibit coincident or overlapping distribution. 



The species of Sagitta, though relatively few-" 

 in number, are often so similar in structure- 

 as to be separated with great difficulty even byr 

 a specialist. One of the most closely related; 

 groups in the genus is that formed by S^- 

 serratodentata, a cosmopolitan eurythermal- 

 species ranging in the epiplankton from tha 

 latitude of the Straits of Magellan to the- 

 equator, and by S. hedoti, S. ferox and 8~- 

 robusta from the epiplankton of the Indo- 

 Pacific region. These four species occuf- 

 together repeatedly in the Siboga collections^ 

 of surface plankton. A second group of re- 

 lated species is formed by 8. enflata, S. hexap— 

 tera and 8. pulchra which likewise occur' 

 together in the same collections often in larget- 

 numbers. The first two species of this sec- 

 tion are ones of wide range and the last aiK 

 Indo-Austral one. 



S. neglecta and *S'. regularis, two closely 

 related neritic species from the Malay Archi- 

 pelago, form a third group having a coincident 

 distribution. There are possibly two in- 

 stances, S. furcata and S. planctonis, and 8.^ 

 hedoti and 8. sihogm, where related species*- 

 have a contiguous distribution in epiplankton^ 

 and mesoplankton. The data are too scanty 

 for definite conclusion. 



It appears from Fowler's records that, as a.. 

 rule, the most closely related species of thift- 

 genus have, not a contiguous, but a coincident. 

 distribution. In the present condition of ouff- 

 knowledge it is impossible to state whether' 

 these species breed at the same seasons or not^ 

 Shipley in the Cambridge Natural History- 

 states that Sagitta breeds throughout the yean- 

 Fertilization is presumably external in many 

 species, though known to be internal int- 

 one and eggs are pelagic, but there are na-- 

 geographical barriers which intervene to pre- 

 vent the swamping out of incipient species*- 

 Provided there are in these related species of 

 Sagitta, no differential breeding seasons, and? 

 no differential levels or temperatures at whicfc 

 extrusion of ova and sperm occur, no assortive- 

 assemblings of the individuals of the given-' 

 species during specialized breeding seasons an^/ 

 no close fertilization, we are compelled to- 

 fall back upon the assumption of a sterility 

 of the cross between an incipient species an(E 



