Dkcembek 21, 1883.] 



SCIENCE. 



793 



adapted for running instead of swimming; and he 

 thus explains the loss of the exopoditesof the conno- 

 pods, the strengthening of theendopodites, the shor- 

 tening of (he abdomen, the loss of power in tlie 

 pleopods, the flatness of the body and abdomen, llic 

 thickening of the integument, and tlie loss of eye- 

 stalks and of the antennary scale. The respiratory 

 function of the pleopods he attributes to the loss of 

 the carapace, and the thickening of the integument. 



The general conclusions of this highly suggestive 

 and interesting paper maybe summarized as follows. 



The Mal.acostraca are descended from the phyllo- 

 pods, among which Nebalia is their nearest relative. 



The Euphausiacoa are the most primitive Malacos- 

 traca. The decapods originated from the Euphau- 

 siacea, although the most primitive decapods, the 

 Natantia, are now widely separated from this ancestral 

 form. The Squillacea stand by themselves, their 

 nearest, although distant, allies being the ]?uphau- 

 siacea. They show in certain points a more primi- 

 tive condition than any other Malacostraca; although, 

 as a whole, they are highly modified. 



The Mysidacea are also derived from the Eupliaii- 

 siacea; although they are so different from them that 

 they must be placed in a distinct order, and the 

 group Schizopoda must be abandoned. The Mysi- 

 dacea have no close relationship to the decapods. 



The Cumacea arise from the Mysidacea, and the 

 amphipods and isopods from a form between the 

 Mysidacea and the Cumacea. The amphipods and 

 isopods are not a primitive group distantly related to 

 the Podophthalmata, but they are the most highly 

 specialized of the Malacostraca. 



lie gives the following as his phylogenetic classifi- 

 cation of the Crustacea : — 



Amphipoda. 



Decap Jil:). 



LETTERS TO THE EDITOR. 

 Radiant heat. 

 Mk. FrTZOKUAi.i) has favored me with a paper ' in 

 which he takes exception to my views respecting 

 radiant heal,-' wherein he says, — 



" Suppose that two regions, A and i?, be separated by three par- 

 allel Bcreeni*, /, m, and u, having aperturet* in them, x, y, z, capa- 



I 'I I' 



ll' ll' I I 



W. K. liRO.»K« 



ble of being opened and clu.suil from the *:umre, 80 ae to make every 

 tiling perfectly syinmetricul round the line .Ifi, perpendicular lo 

 the screens. Now, if x be opened for a very short time, a certain 

 quantity of radiant energy will becape out of .1 into the region 

 between I and m; and if y be opened when this heat reache8 rn. 

 it can certainly be let on into the region mn; and if s be similarly 

 opened when it reaches it, this radiant heat will get into B. 

 While 3 was open, however, some heat left Ji; but, as Dr. Kddy 

 observes, y may be closed so as not to let this even get througli 

 the screen m, and it can be all returned into B by reflection 

 through s or some other aperture. So far I entirely agree with 

 Dr. Eddy, and so far it seems an if the rcBult bad been to trans- 

 fer heat from A to />' without J?'b lof*hig any heat by having it 

 transferred to A. As I warned Dr. Eddy when I heard his 

 paper, there arc, however, other bodies and regions to be con- 

 sidered besides A and H. There are more than two bodies con- 

 sidered : there is the region of the screens. Consider what hap- 

 pens when the heat that escaped out of 5 into the nxn region tries 

 to get baciv into B. Some door must be opened lo let it pass; 

 and, while it is passing in, an at least equal amount of heat will 

 be passing out of B into the tun region, po that you can never 

 really get the heat that has once left B back into £ again. This 

 is true, whether you adopt doors over fixed apertures, such as I 

 have supposed, or moving apertures, such as Dr. Kddy proposed. 

 What really takes place is this: a certain quantity of heat es- 

 capes out of A and reaches B ; and a not less quantity of heat 

 leaves B, and is kept entangled in the region of the screens, and 

 it is only possible to let the heat pass from A to B by means of 

 this third region. Hence this only really comes to the same 

 thing as letting A radiate some of its beat into the screen region, 

 while B is kept closely shut up. Xow, be it ob8er\*ed that Dr. 

 Eddy practically postulates that this screen region is at least 

 colder than -■! — in fact, he assumes it to be perfectly cold, I.e. to 

 contain no radiant heat except what is admitted from A and B^ 

 so that it is by no means contrary to the theoi^ of exchanges that 

 A might cool by radiating into this region." 



Now, Mr. Fitzgerald has stated only two of the 

 three things whicli occur while the door z is open. 

 He omits lo state, that in my process a certain amoimt 

 of heat whicli has come from A also passes through 

 the door z every lime it is opened, into the region li; 

 and this is alt which is proposed to be accomplished 

 by the process which is at all itnusual or peculiar. 

 Thus the fact remains, thai although a definite amount 

 of heat from B remains entangled in the region miXy 

 which is not increased with the lapse of time, there 

 is a continued passage of heat through this region 

 into By that being the very object sought to be accom- 

 plished by my process. It is not easily seen how 

 the arrangement of screens and apertures proposed 

 by Mr. Fitzgerald could be so manipulated as to 

 bring the heat coming from A into a position such 



' On Dr. Kddy*s hypothesis that radiant heat Is an exception 

 to the second law of thermodynamics. By George F. Fitzger- 

 ald, M.A., F.T.C.D., He. proc. roy. Duhl. «oc., iv. pt. I. 



» A'c. proc. Ohio mech. ingt., July, 1882. 



