An the neural arches in both M. HS-173 manufacturer Cynaroside web pelikani and H. longicostatum ([3], character 125; [5], character 158). Also, based on the presence of haemal arches in the tail, the centrum is considered to be the pleurocentrum ([13]; [6], characters 310?12). As in M. pelikani, trunk intercentra are absent in H. longicostatum ([6], character 316). The tail of M. pelikani is foreshortened relative to trunk length; however in H. longicostatum the tail is qhw.v5i4.5120 too incompletely known to score character 162 of Huttenlocker et al. [5] as anything other than `?’, even though, based on partial tails, the region probably is foreshortened as well. Pectoral and Pelvic Girdles. The present study provides new information on the interclavicle of H. longicostatum, which was not described fully in the past. In the analysis by Anderson [3], character 150 (see also [5], character 186) was not scored for that taxon because it was unknown if the interclavicle was diamond- or T-shaped. Although I did not observe a complete stem in H. longicostatum, the overall shape of the interclavicle is the same as that of M. pelikani, which implies that the element is T-shaped when complete. However, character 240 of Ruta and Coates [6] should remain scored as “?” for H. longicostatum because the exact nature of the stem is unknown. As in M. pelikani, the interclavicle is wider than long ([6], character 241). Also, the first record of a partially ossified coracoid in M. pelikani affects previous scoring for character 159 of Anderson [3] and character 198 of Huttenlocker et al. [5], which coded the element as `unossified.’ The number of coracoid foramina is unknown in both M. pelikani and H. longicostatum; both pnas.1408988111 taxa should be scored as `?’ for character 158 of Anderson [3] and character 197 of Huttenlocker et al. [5]. Finally, in the pelvic girdle of a large individual of H. longicostatum, the ilium is slightly bifurcated at the dorsal tip rather than terminating as a single blade ([3], character 170; [5], character 211; [6], character 272). The state of that character is dependent upon the maturity of the specimen because less-developed individuals do not exhibit bifurcation. Limbs. Previously, the entepicondylar foramen was thought to be absent in H. longicostatum ([3], character 160; [5], character 199; [6], character 253; [47], character 7). However, I observed the foramen in a single individual. Other specimens of H. longicostatum are not preserved in an orientation in which the foramen would be visible, although ontogenetic variation cannot be excluded. The amount of torsion present in the humerus is affected by the degree of morphogenesis in both H. longicostatum and M. pelikani ([3], Character 161; [5], character 200). In developmentally younger individuals, the torsion may be as low as 0?0? However, in mature specimens the torsion approaches 90? The prominence of the deltopectoral crest ([3], character 162; [5], character 201) also is subject to ontogenetic variation. In both taxa, the crestPLOS ONE | DOI:10.1371/journal.pone.0128333 June 17,57 /Skeletal Morphogenesis of Microbrachis and Hyloplesionis absent in the least mature specimens. The structure becomes distinct in large individuals, although it never reaches the relative size or degree of prominence found in other lepospondyls like Pantylus, and many large temnospondyls. A distinct ectepicondylar ridge is not developed in either M. pelikani or H. longicostatum ([6], character 255, previously unscored for M. pelikani). Presen.An the neural arches in both M. pelikani and H. longicostatum ([3], character 125; [5], character 158). Also, based on the presence of haemal arches in the tail, the centrum is considered to be the pleurocentrum ([13]; [6], characters 310?12). As in M. pelikani, trunk intercentra are absent in H. longicostatum ([6], character 316). The tail of M. pelikani is foreshortened relative to trunk length; however in H. longicostatum the tail is qhw.v5i4.5120 too incompletely known to score character 162 of Huttenlocker et al. [5] as anything other than `?’, even though, based on partial tails, the region probably is foreshortened as well. Pectoral and Pelvic Girdles. The present study provides new information on the interclavicle of H. longicostatum, which was not described fully in the past. In the analysis by Anderson [3], character 150 (see also [5], character 186) was not scored for that taxon because it was unknown if the interclavicle was diamond- or T-shaped. Although I did not observe a complete stem in H. longicostatum, the overall shape of the interclavicle is the same as that of M. pelikani, which implies that the element is T-shaped when complete. However, character 240 of Ruta and Coates [6] should remain scored as “?” for H. longicostatum because the exact nature of the stem is unknown. As in M. pelikani, the interclavicle is wider than long ([6], character 241). Also, the first record of a partially ossified coracoid in M. pelikani affects previous scoring for character 159 of Anderson [3] and character 198 of Huttenlocker et al. [5], which coded the element as `unossified.’ The number of coracoid foramina is unknown in both M. pelikani and H. longicostatum; both pnas.1408988111 taxa should be scored as `?’ for character 158 of Anderson [3] and character 197 of Huttenlocker et al. [5]. Finally, in the pelvic girdle of a large individual of H. longicostatum, the ilium is slightly bifurcated at the dorsal tip rather than terminating as a single blade ([3], character 170; [5], character 211; [6], character 272). The state of that character is dependent upon the maturity of the specimen because less-developed individuals do not exhibit bifurcation. Limbs. Previously, the entepicondylar foramen was thought to be absent in H. longicostatum ([3], character 160; [5], character 199; [6], character 253; [47], character 7). However, I observed the foramen in a single individual. Other specimens of H. longicostatum are not preserved in an orientation in which the foramen would be visible, although ontogenetic variation cannot be excluded. The amount of torsion present in the humerus is affected by the degree of morphogenesis in both H. longicostatum and M. pelikani ([3], Character 161; [5], character 200). In developmentally younger individuals, the torsion may be as low as 0?0? However, in mature specimens the torsion approaches 90? The prominence of the deltopectoral crest ([3], character 162; [5], character 201) also is subject to ontogenetic variation. In both taxa, the crestPLOS ONE | DOI:10.1371/journal.pone.0128333 June 17,57 /Skeletal Morphogenesis of Microbrachis and Hyloplesionis absent in the least mature specimens. The structure becomes distinct in large individuals, although it never reaches the relative size or degree of prominence found in other lepospondyls like Pantylus, and many large temnospondyls. A distinct ectepicondylar ridge is not developed in either M. pelikani or H. longicostatum ([6], character 255, previously unscored for M. pelikani). Presen.