March 24, 1881] 



NA TURE 



4^3 



the sea-water to a depth of several feet. Samples for examina- 

 tion were obtained by " dipping" with a bucket, as well as by 

 the tow-net. It seemed to be a Confervoid Alga. 



On slightly agitating the ^^ater in a glass jar, the fluffy masses 

 broke up into minute particles, which, under a magnifying power 

 of sixty diameters, were found to be composed of spindle-shaped 

 bvindles of filaments. Under a power of 500 diameters these 

 filaments were seen to be straight or slightly-curved rods, articu- 

 lated but not branching, and divided by transverse septa into 

 cylindrical cells, which contained irregularly-shaped masses of 

 granular matter. These rods, which seemed to represent the 

 adult plant, measured ^nrVij inch in width. On carefully examin- 

 ing many samples, some filaments were detected, portions of 

 which seemed to have undergone a sort of varicose enlargement, 

 being more than twice as nide as the normal filaments. These 

 propagating filaments (if I am right in so calling them) were 

 invested by a delicate tubular membrane, and contained some 

 granular semi-transparent matter, in which was imbedded a row of 

 discoid bodies ; the latter appearing as if about to be discharged 

 from the ruptured extremity of the tube. These bodies measured 

 j-jijijth of an inch in diameter : when seen edgewise presented a 



lozenge-. --haptd appearance, ard were devoid of cilia or striae. 

 Conjugation was not obFcrved. 



On allowing a jarful of the sea- water to stand by for twenty- 

 four hours it was found that the confervoid matter had all risen 

 to the surface, forming a thick scum of a dull green colour, 

 while the water had assumed a pale pui-ple colour, resembling 

 the tint exhibited in a weak solution of permanganate of 

 potash. 



From November 24 to 29, during which time the 'ship tra- 

 versed slowly a distance of 330 miles, the sea contained these 

 organisms. For the first three days the large patches were 

 frequently in sight, and during the rest of the time the sur- 

 rounding water presented a dusty appearance from the presence 

 of the tiny .spindle-shaped bundles. On the evening of the 26th 

 an unusually dense patch was sighted and mistaken for a reef, 

 being reported as such by the look-out-man aloft. 



Sydney, January 24 R. W. Cofpinger 



Feeding a GuU with Corn 



In Prof. Semper's recently-published work on the "Condi- 

 tions of Existence as they Affect Animal Life," a review of 

 which from the pen of Prof. Lank ester appeared in your columns 

 a fortnight ago (vol. xxiii. p. 405), allusion is made on pp. 67, 68, 

 and elsewhere to John Hunter's celebrated experiment of feeding a 

 gull with corn. Prof. Semper, however, seems not to have 

 been aware of the precise nature of the result of Hunter's experi- 

 ment. He says: "The English anatomist Hunter purposely 

 fed a sea-gull for a whole year on grain, and he thus succeeded 

 in so completely hardening the inner coat of the bird's stomach. 



which is naturally soft and adapted to a hsh diet, that in appear- 

 ance and structure it precisely resembled the hard, horny skin of 

 the gizzard of a pigeon." 



The original account, I believe, of Hunter's e.xperiment, was 

 published in Sir Everard Home's "Lectures on Comparative 

 Anatomy" (vol. i. p. 271, 1814), and an extract from that work 

 is appended to the description of Hunter's original preparation, 

 still preserved in the College of Surgeon*, in the descriptive 

 catalogue of that collection (vol. v., 1S33, pp. 149-50, Prep. 

 523). What Hunter succeeded in effecting was to very much 

 increase the thickness of the musciilai- walls of the gizzard, 

 which, as may be seen by comparing his specimen (No. 523) 

 with that of the stomach of another gull close by, have become 

 developed to an extent about double their usual size. There is 

 no manifest increase in the thickness of the "inner" — or so- 

 called "epithelial" — coat of the stomach visible in the prepara- 

 tion, nor do Home or Owen allude to any such feature in their 

 descriptions. Hunter's experiment, therefore, simply comes 

 under the numerous well-ascertained instances of the increased 

 development, consequent on increased use, of muscle, and has 

 no real connection with the "modifying effects of food," such as 

 that produced in canaries by feeding them on cayenne pepper, 

 and others cited by Semper. W. A. Forbes 



Zoological Society's Gardens, N.W., March iS 



Dynamics of "Radiant Matter" 



As the chief object of Mr. Preston's paper under the above 

 title in N.\ture, vok xxiii. p. 461, seems to be to support Le 

 Sage's "shelter theory" for gravity, you will perhaps let me 

 point out one objection to that theory in any form which has 

 hitherto been deemed conclusive, and with which Mr. Preston 

 does not deal. It is that under it gravity would not vary, as it 

 is known to do, equally with mass, but would vary net equally. 



The theory applies perhaps so long as you consider only the 

 case of isolated atoms, but it fails entirely when applied to 

 clusters of atoms. 



Observation show s that gravity varies only with distance and 

 with mass ; but if it were caused by any form of shelter hitherto 

 imagined, it would vary also with density and with bulk in such 

 a way that a pound of, say water, would weigh more than a 

 pound if raised into steam, because its atoms, in loose order as 

 steam, would give each other less shelter from the action of the 

 kinetic .^ther than when in close order as water, and in such a 

 way also that two spheres of, say iron, each weighing one poimd, 

 would weigh less than two pounds if welded into one sphere, 

 because some atoms in the one sphere would be better sheltered 

 than any atoms in the two .spheres. \Vm. Muik 



March 21 



The Oldest Fossil Insects 



Mr. S. H. Scudder has published (Anniversary Memoirs of 

 the Boston Society of Natural History, iSSo, pp. 41, i:late l) a 

 memoir on the Devonian Insects of New Brunswick. The 

 fragments of the six described species were discovered by the late 

 Prof. C. F. Harttin 1S62, and have been since 1S65 described in 

 several papers by the same author. The new paper is a very 

 detailed and elaborate one, with entirely new and improved 

 figures, and is followed by a number of conclusions, as the final 

 result of his work (Report, Anicr. /oiir/i. of Sci., Feb. 1S81). 

 The conclusions would be of prominent importance for the 

 history of the evolution of insects, if they could be accepted 

 without reserve. Of course facts and conclusions should 

 be able to stand the most severe test ; and that is not the 

 case with this publication. "As the simpler Devonian insects 

 have certain special relations," he says, "with the Ephemeridse, 

 their description is preceded by an account of the wing-structure 

 of the modem Mayflies as a basis of comparison " (p. 4). 



The simple fact that not one of the described species has any 

 relationship to the Ephemeridce is sufficient to cause us to object 

 to his descriptions and conclusions related to this family. This 

 statement is not based upon a difference of opinion, but simply on 

 the evidence of facts which cannot be denied by any one conver- 

 sant with the families Ephemeridce and Odonata. 



Platcphcniera antiijua is a part of the apical half of the wing, 

 without the tip, of a gigantic dragonfly. The suddenly naiTowed 

 secc nd cubital space is to be found in Isophlebia of the Solenhofen 

 slate. The imperfectness of the fragment allows no further con- 

 clusions. 



Ccrephoiura simplex is a diagonal fragment of the middle ot 



