536 



SCIEJSrCE. 



[Vol. IX., No. 226 



tion of the narrow bands formed by the marine 

 stages in Mississippi and Alabama ; with a manifest 

 north-westward trend of such deposits as are con- 

 tinnonsly traceable in north-w^estern Louisiana, 

 while the later stages are abruptly deflected to the 

 sonth-west, — all points to a rapidly progressing 

 elevation of the axial cretaceous trough, that may, 

 or may not, have completely separated the interior 

 from the Gulf waters before the beginning of the 

 tertiarj'^ period. In any event, the region referred 

 to appears to me to be a critical one, deserving of 

 exhaustive examination in advance of many others 

 that offer only a subordinate interest in comparison 

 to the problem of the correlation of the intraconti- 

 nental and the marine tertiary. E. W. Hilgabd. 

 Berkeley, Cal., May 17. 



The pelvis of the dugong. 



As far as I am aware, the pelvis of Halicore aiis- 

 tralis has never been projjerly described or figured. 



Last fall I had the opportunity of examining, here 

 at my father's establishment, six ligamentary skele- 

 tons, embracing both sexes, of this animal. A few 

 hasty notes made at the time, and a section of verte- 

 brae, including the pelvis (in which, ixnfortunately, 

 the ischia have been torn asunder and separated 

 from their haemapophysis), is all the material I can 

 laj^ hands on, now that I have time to look the mat- 

 ter up : consequently my drawing and description 

 cannot include a few points that I would wish. 



In all six cases the fourth post-dorsal vertebra is 

 the first sacral. The ilia are connected to the distal 



ends of its diapophyses by short ligaments. The 

 ends of these diapophyses are greatly swollen dorso- 

 ventrally, their vertical diameter being thirty-three 

 millimetres, whereas the preceding one measures 

 but ten millimetres. The diapophyses of the two 

 succeeding (sacral) vertebrae are also decidedly 

 thicker at the ends than is the case in either the last 

 lumbar or the succeeding caudals. Anchylosed to 

 the ilia are the ischia lying in the same line, and 

 showing their junction by a prominent swelling in 

 the mass of the bone. 



The distal ends of the ilia were connected with 

 each other by a short ligament, and separated from 

 the apex of the haemapophysis of the second suc- 

 ceeding vertebrae bj'^ but a few millimetres, con- 

 nected to it either by a ligament or muscle, but 

 which it is now too late to determine. 



The ilium is 109 millimetres in length ; the ischi- 

 um, 102 ; the transverse diameter of its distal end, 

 46 ; the anterior-posterior length of the symphysis 

 ischia, 34. 



The first haemapophysis consists of two straight 

 V-shaped bones 30 millimetres long, 29 millimetres 



apart at bases, with points diverging to a distance of 

 51 millimetres. The next, to which the ischia join, 

 has its two parts curving inward, leaving an oval 

 opening, the extremities not quite meeting, and lig- 

 amentously connected. The succeeding haemapophy- 

 ses have their ends anchylosed, and are V-shaped. 



The j)oint that I especially wish to emphasize is, 

 that the pelvis is not vertical to the axis of the verte- 

 bral column, but lies at practically the same angle as 

 ordinarily obtains in the mammalia. 



In the six specimens examined, two had nineteen 

 thoracic vertebrae, while four had but eighteen. All 

 had three lumbar vertebrae. The thoracic are gen- 

 erally stated as being nineteen in number : with 

 these this was the exception. 



It is further to be noticed that the dugong appears 

 to be an exception to the rule that when the number 

 of thoracic vertebrae is increased or diminished 

 there is a compensating diminution or increase in the 

 number of lumbar vertebrae. Henbx L. Wakd. 



Rocliester, N.Y., May 34. 



A cretaceous river-bed. 



The springs at San Marcos, Hays county, Tex., 

 where the San Marcos Eiver rises full grown from 

 the earth, with a steadiness of flow in marked con- 

 trast with the majority of Texas rivers, are, aside 

 from their scientific aspects, sufliciently interesting 

 to have been a subject of popular speculation and 

 newspajDcr discussion ever since the settlement of 

 Texas. The theories that have been advanced are 

 various, from the popular idea that it is sufficiently 

 explained by the presence of a cave full of water 

 under the hill, to the explanation proposed by an 

 imaginative newspaper editor, that the water comes 

 underground from the Eocky Mountains. 



I have not felt it necessary to familiarize myself 

 with the details of this discussion, since, although 

 my conclusions may be to some extent old, the proof 

 is certainly new; for the general principle upon 

 which it is based has been but recently announced 

 by Mr. Eobert T. Hill in the American journal of 

 science for April (xxxiii. p. 29) ; namely, that there 

 exists between the earlier cretaceous strata of Texas 

 and the superimj)osed rocks a plane of ' non-con- 

 formity by erosion,' indicating an interval of emer- 

 gence between the two i^eriods of cretaceous rock 

 formation. 



The strata in the vicinity of San Marcos not only 

 furnish a striking proof of the trixth of this princi- 

 ple, but thej^ become a key to whatever is mysterious 

 in the origin of the San Marcos Eiver. 



The accompanying section roxrghly represents the 

 rocks exposed by the San Marcos at its source. 



No better stratigraphical landmark than the stra- 

 tum 66, the Exogyra arietina marl, could be desired. 

 The exposures at San Marcos are typical ones, con- 

 taining an unusually large proportion of perfect 

 bivalve specimens of Exogyra arietina E., besides 

 the usual smaller quantity of GrypharaPitclieri, etc. 

 Its exposures are from fifty to one hundred feet 

 above the river-level, and, in connection with the 

 Ostrea carinata bed below, furnish conclusive proof 

 that these rocks are of the Washita division of the 

 earlier or Texas cretaceous; lacking, however, the 

 uppermost members of that series. 



In the little valleys back of the portion of the 

 section marked aa, I found a conglomerate composed 

 of fragments of the hard earlier limestones and 



