Two datasets are provided to support the journal article (https://doi.org/10.1643/i2020112) by T. C. Grande and M. V. H. Wilson naming the new Bermuda fish species Polymixia hollisterae. The first dataset is for 2-D multivariate morphometric comparisons of selected specimens and species of the fish genus Polymixia. The file is in TPS format, as a plain text file, for use in the application MorphoJ. The data are for 27 specimens with pixel coordinates for 34 landmarks digitized in ImageJ and used to generate Fig. 13 in the referenced publication. The second dataset, published here courtesy of Dr. R. Eytan, is a mitochondrial DNA barcode sequence in fasta format for the second paratype specimen, a small juvenile of the new species. The specimen is only the third one known of the new species and the only one not from Bermuda. It was collected in the north-central Gulf of Mexico and is now deposited in the Harvard University MCZ fish collection as catalog number MCZ 174218. The fasta file can be used as input to the barcode identification function at boldsystems.org, although its original identification in BOLD was incorrect.

The multivariate morphometric data were obtained using the PointPicker plugin for ImageJ, using images of each of 27 specimens. Data for each specimen occupy 36 lines, the first one giving the number of landmarks and the last one representing the specimen identifier.  Each of the other 34 lines represents the coordinates of a landmark - an X-coordinate and a Y-coordinate in image pixels, with the X-values counting from the left and the Y-values counting from the top of the image.

There are no missing values in the morphometric data. The specimen identifier is the last line of each specimen's entry in the tps text file for the morphometric analysis. The barcode sequence is in fasta format as a plain text file. The contents of the file can be pasted as an unknown into the identification function at boldsystems.org.