596 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.xxv. 



cli.stingui.sh the variety magna. AH of the characters are full.v covered 

 by the variations in the same characters in ^1. mgittallii and its A'ariety 

 transvcrm. The nearly circular form of variety magna is about the 

 onh' character that can be considered of value, and that is ver^- closely 

 approached by some specimens of A. saglttalis. 



Dr. ^latthew's making his shell a variety of the variety ht<Jt! {Acro- 

 treta mgittaHs helti) is not altogether desirable, and it is not probable 

 that the Lower Tremadoc shell IjcU'i is a variety of ^1. saglttall.s, and 

 still less so that the New Brunswick shell is a variet}' of Acrotreta 

 helti In order to have all the data possible for the student, I have 

 illustrated Dr. Matthew's types, which he kindly sent me for the 

 purpose: also specimens I found at Hastings Cove. 



lun'tiiiiiioit and locality. — Middle Caml)rian, Pai'ado,r<de>< zone, 

 Hastings C'ove. Kennebecasis River. St. John County, New Brunswick. 



ACROTRETA SAGITTALIS TRANSVERSA Hartt. 



Ohoh'Ua tvansrevm Hartt, Dawson, 1868; Acadian Geology, 2(1 e<l., p. 614. — 

 - Walcott, Bull. U. S. Geol. Sur., No. 10, p. 16, 1884, pi. i, fig. 5. 



Linnarssoida tvanfurrsa Walcott, Am. Jonrn. Sci., 3d ser., XXIX, 1885, ]>. IK), 

 %s. 3, 46.— :\l.\TTHEW, Trans. Roy. Soc. Can., Ill, 1886, p. 35, pi. v, fig. 11.— 

 Hall and Clarke, Pal. X. Y., VIII, 1892, Pt. 1, p. 108, pi. iii, figs. 3S-4i'.— 

 Mattheav, Trans. X. Y. Acad. Sci., XIV, 1895, ji. 125, pi. v, figs. 1, 2. 



Liiinarssonia sagiUaliH var. Iransrersa Walcott, Tentli Ann. Rept., V. R. i'-e-n. 

 Sur., 1891, description of pi. lxviii, fig. 2a. 



This is the representative of .1. sagtttalis in the I*aradoxide.s zone of 

 New Brunswick. Many shells are more transverse in outline than the 

 average of ^1. sagittaUs^ but examples of the latter are nearly as nuuh 

 so, and the interior casts show a striking similarity in the Welsh and 

 Acadian forms. The range of variation in each form is nearlv if not 

 quite as great as "the variation between the two. In view of this, I 

 think it ])est to characterize transversa as a variety of A. sagittalls. 



Fortnation and locality. — Middle Cambrian, Paradoxides zone, 

 division b, c^ and d, St. John. Hanford Brook, New Brunswick; 

 Paradoxides dari.si zone. Manuels Brook section. Avalon Peninsula. 

 Newfoundland. 



ACROTRETA SAGITTALIS TACONICA Walcott. 



Linnarssunia iaconica Walcott, Am. .Tour. 8ci., 3<1 ser., XXXIV, 1887, p. 189, 



pi. I, figs. 18, 18a-d. 

 Linnarsmnia sagiUaUs \ar. taconira Walcott, Am. Jour. Sci., 3d ser., XXXVIII, 



1889, p. 36; 1891; Tenth Ann. Rept. U. S. Geol. Sur., p. 610, pi. lxviii, 



1, la-d. 



This variety is characterized ])y the uniformly shorter median ridge 

 of the dorsal valve. The pedicle valves appear to be identical with 

 those of A. sagittalis from Andrarum, which are embedded in the 

 same character of matrix as the A'ariety taconica. 



This is the second species found in association with Olenellus. A. 



