BERYCOID FISH FAMILY POLYMIXIIDAE LACHNER 191 



Chace, division of marine invertebrates, U. S. National Museum, 

 presented me with considerable literature on the distribution of 

 marine invertebrates in the Atlantic Ocean. 



The method of recording counts and measurements of several 

 characters require explanation because they present certain specific 

 or unique problems in this group. The last ray of the soft dorsal 

 FIN is nearly divided to its base and may be erroneously interpreted 

 as two. One must be especially cautious when the fins are badly 

 frayed and torn. All rays of the pectoral fin were counted. The 

 VERTICAL SCALE ROWS Were countcd by noting the oblique rows on 

 the body above the lateral line, commencing with the row at the 

 junction of the lateral line and the gill opening, and countmg pos- 

 teriorly to the end of the hypural plate. This count was less variable 

 than counting the scales along the lateral line and can be made with 

 greater accurac3^ The lateral line pores were counted from the 

 upper edge of the gill opening to the end of the hypural plate. 

 The total number of gill rakers recorded includes all developed 

 rakers on the upper and lower limbs of the fu-st right arch. Several 

 depressed, usually elongate rudiments occur adjacent to the last, 

 small, pointed raker on the lower arch. These were omitted from 

 the count for two or more may be coalesced. A small tubercle may 

 also occur adjacent to the last, small, raker on the lower arch. When 

 this tubercle was higher than the diameter of its base it was con- 

 sidered as a raker and included in the count, and when smaller it 

 was considered as a rudiment. There is a reduction in the number 

 of gill rakers with increase in body size, giving rise to the negatively 

 skewed curve as shown by the frequency distribution of the total 

 number of gill rakers of lowei from Cuba in table 2. This is due to the 

 presence of a larger number of small specimens. The positively skewed 

 curve shown for the same species from the Gulf of Mexico was caused 

 by the great number of large specimens. The longest measurement of 

 the UPPER JAW was taken because the jaw has a cleft at the symphysis. 

 The LENGTH OF THE PECTORAL FIN was measured from the tip of the 

 longest ray to the anteriormost portion of its base. All measure- 

 ments of the LENGTH OF A FISH refer to the standard length. In the 

 numerical data given in the description, the mean is listed first, and 

 it is followed by the range of variation for each character showing 

 interspecific differences. 



Diagnostic characters 



The species have differentiated almost entirely in meristic charac- 

 ters. The data are tabulated in the form of frequency distributions 

 in tables 1 to 6, and in most cases are segregated by localities. 



