748 



SCIENCE 



[N. S. Vol. XXXIII. No. 854 



The fact that all of the mail-cheeked fishes 

 do not have the coracoid elements separated 

 by the actinosts is not an argument in dis- 

 favor of the relationship of the Gobiidae to the 

 Cottidse, because the mail-cheeked fishes with 

 the typical shoulder girdle (such as the Scor- 

 psenidee) vcere, of course, the ancestors of the 

 Cottidse. From the Cottidse came the Lipa- 

 ridse and the Cyclopteryidse, as Dr. Gill long 

 ago pointed out.' 



It does not seem improbable that the gobies 

 may have come from some ancestor — probably 

 scale-covered — of the Cottidffi in which the 

 shoulder girdle had become differentiated. 

 Further, it is not altogether improbable that 

 this ancestor might also have been from some- 

 where along the line leading towards the 

 Cyclopteryidffi and the Liparidse; some form 

 in which the ventrals had just become at- 

 tached to each other, much as in most of the 

 gobies of to-day. From this the sucking disk 

 of the Liparidse and Cyclopteryidse could have 

 developed. In considering this supposition, 

 of course, we could only explain the gobies 

 with separate ventrals by the separation being 

 secondary. The gobies further resemble the 

 last two families in having no myodome to the 

 cranium. 



It is conceded, certainly, that the family 

 Gobiidas is not very close to the Cottidse, they 

 having lost the suborbital stay to the pre- 

 operculum and undergone other changes, and 

 no modification of the suborders containing 

 these two families is suggested. The char- 

 acter of the shoulder girdle seems to be the 

 most significant character in showing a pos- 

 sible line of descent of the gobies, and it is 

 suggested in light of it that the group be 

 placed in close relationship with the mail- 

 cheeked fishes in works involving classifica- 

 tion. With this question in mind the gobies 

 should, of course, be studied in detail. 



Edwin Chapin Staeks 



carcharias borneensis and baebus elongatus, 

 as preoccupied names 



In the Philippine Journal of Science, Vol. 

 V, No. 4, Section D, October, 1910, p. 263, 



' Froc. V. S. Nat. Mus., Vol. XIII., 1890. 



PL 1, Mr. Alvin Scale describes, as new, 

 " Charcharias horneensis." This is preoccu- 

 pied by Carcharias (Prionodon) horneensis 

 Bleeker, Act. Soc. Sci. Ind.-Neerl. (Borneo 

 12), V, 1858-59, p. 8. 



In the same journal Mr. Scale also de- 

 scribes, as new, Barhus elongatus, on p. 265, 

 illustrated on PI. 2 as Fig. 1. This is preoc- 

 cupied by Barhus elongatus EiippeU, Mus. 

 Senckenb., II, 1837, p. 11, PI. 2, Fig. 1. 



Henry W. Fowler 



ACADEMT OF NATTJBAL SCIENCES, 



Philadelphia, Pa., 

 February 9, 1911 



SOCIETIES AND ACADEMIES 



THE PHILOSOPHICAL SOCIETY OP WASHINGTON 



The 693d meeting was held on April 8, 1911, 

 President Day in the chair. Three papers were 

 read. 



Mechanical Forces on an Electrical Conductor: 

 Dr. Feank Wenner, of the Bureau of Stand- 

 ards. 



Starting with the equation for the electromotive 

 forces in an inductive circuit the speaker showed 

 how it follows under certain conditions that a 

 current through a conductor causes forces such as 

 to require a tension in the conductor to maintain 

 equilibrium, that is, the forces tend to increase 

 the length of the conductor. It also follows that 

 under other conditions the forces are such as to 

 tend to decrease the length of the conductor. 

 Under most conditions the force on an element of 

 the conductor near the surface is such as to tend 

 to crowd it toward the center. 



It was also stated that it is possible that a 

 current in a conductor causes forces other than 

 those due to electro-magnetie and electrostatic 

 actions, the former only having been considered 

 by the speaker. So far, however, no one has 

 shown the presence of any such additional force. 



The Completion of the Texas-California Arc of 



Primary Iriangulation : Mr. Wm. Bowie, of the 



Coast and Geodetic Survey. 



Three grades of triangulation are recognized: 

 primary, secondary and tertiary; and the grade 

 depends upon the accuracy of the angle and length 

 measurements rather than upon the length of line 

 between pairs of stations. 



The primary work is extended in long arcs over 



