3 I I IM II. I IN 82, UNITED STATES NATIONAL MUSEUM 



Carpenter -aid that the special character (hat distinguishes tin* Milberti group is 

 tin' large Bize of one or snore of the Becond, third, and fourth pairs of pinnules, and 

 remarked that thia feature is well ahown in atuepa ( Rettrometra qwndupliaava, 

 LO-armed individuals), rnUberH (=tessellata, or more likely, molleri), variipinna 

 | Uelerometm en nulata, LO-armed individuals), and also in some species described by 

 Bell in the Alert report, while it reappears in a more marked degree in some of the 

 multibrachiate apeoies. He noted that in some cases, aa in milberti (referring chiefly 

 to molleri), the Becond, thud, and fourth pairs are all of greater size than the pinnules 

 above and below them, sometimes the second and sometimes the thin! pairs being 

 the largest. He remarked that in anceps (=10-armed individuals of ll,t,mmetra 

 guindupHcaoa) and in variipinna ( LO-armed individuals ot Heteromelra emvufata) 

 the pinnules of (lie fifth or sixth brachials (P b and P 3 ), or both, are considerably longer 

 and stouter than their fellows, and in carjn nti / i ( = Oligometra curjx nil ri) and in jiumifn 

 (=Camp»ometra loveni) the large pinnule is on the fourth brachial (P 2 ). But in 

 carinaia (= Tropiometra carinata) and in pariicirra (=Dorometra parvicirra) the 

 third and following pairs of pinnules are much more equal in size. 



At the end of the MHb< rti group Carpenter placed two "abnormal" species, 



■wiM i Decametra informis) and loveni ( = Colobometra perspinosa), in which P a 

 is absent, though in other respects, be said, they conform pretty well to the general 

 type. 



In my lirst revision of the genus Antedon published in 1907 the species placed by 

 Carpenter in the Milb, rti group were distributed in the following genera — Himero- 

 meira [milberti, tesaeUata,, anceps, and variipinna); Antedon (serripinna, carpenteri, 

 pumila, and parvicirra); Cyllometra [perspinosa, informis, and loveni [renamed Cyllo- 

 metra belli]); and Tropiometra (fiorinata); pinnijormis and laerissima were not men- 

 tioned. 



Tow aid the end of 1908 I described Himerometra molleri, H. schlegelii, H. producta, 

 and //. i nsiftr from specimens in the collection of the Zoological Museum at Copen- 

 hagen, and about two weeks later I described //. discoidea from specimens in the 

 United States National Museum. 



In the revision of the family Himerometridae published in 1909 I established the 

 new genua Amphianetra with Comatvla (Alecto) milberti J. Muller, 184G, as the geno- 

 type, including in it unrips, ensiformis (sic), laerissima, milberti, molleri, producta, 

 schlegelii, tessellata, and sariipinna. 



In the kej t" the genera of Himerometridae Amphimetra was paired with Himero- 

 metra, these two genera being differentiated from all the related genera by having 

 the middle and distal brachials extremely short and oblong, and the elements of the 

 [Br Beries and lower brachials strongly convex dorsoventrally, appearing swollen. 

 Prom Hi/men metra it was separated by having the IHr series and first two brachials 

 in apposition for their entire length, strongly developed synarthrial tubercles, P, 

 smaller and slenderer than P 2 , and 10-20 arms, whereas in Himerometra the division 

 Beries are rounded dorsaily and widely separated laterally, synarthrial tubercles are 

 not developed, P| resembles l' D and P p and is larger than I'.., and there are more than 

 25 arms. 



Later in 1909 I described Amphimetra mortenseni (June 19) and A. par ilk (August 



