SCIENCE-GOSSIP. 



273 



all. But nevertheless, anyone breeding the nymphs 

 of the larger dragon-flies could not do better than 

 try Gammarus fluviaiilis first. 



Crawling along the surface of the mud or stones 

 at the bottom of the water, the motion of these crea- 

 tures is sluggish enough, 

 but in case of need they 

 have another means of 

 locomotion. Breathing is 

 carried on by extracting 

 from it the oxygen dis- 

 dissolved in water, and 

 expelling the exhausted 

 liquid rythmically from 

 the aperture at the pos- 

 the body 



the five 



forcible 



insect is 



terior end of 

 surrounded by 

 spines. By a 

 expulsion, the 

 enabled to move rapidly 

 through the water by a 

 series of jerks. As an 

 instance of the force with 

 which the water can be 

 expelled, I may mention 

 that on one occasion a 

 spot was ejected two feet 

 from the receptacle in 

 which the insects were 

 living. The legs are used 

 to assist the forward 

 motion somewhat as a 

 frog's are in swimming, 

 being kept close to the 

 sides during the progress 



of each dart forward and spread out at the com- 

 mencement of the next. On one occasion I noticed 

 a nymph swimming on its back. 



Of my six specimens, one died when on the 

 point of emergence, two were killed and partly 

 eaten by their fellows, one of them just after 



Larva-nymph of Brachytron 



Pratense (about twice 



nat. size). 



moulting, while its body was soft and unprotected, 

 a small one was preserved in spirit for future refer- 

 ence, and two emerged, a male on May 19th, and a 

 female May 24th. The latter, in the nymph state, 

 had been deficient in partof its right middle leg, and 

 the imago also appeared 

 with but half that useful 

 appendage. 



In colour, the larva- 

 nymphs of Brachytron 

 pratense are a fairly uni- 

 fom dark sepia-brown, 

 though slightly mottled 

 in places. Segments six 

 to nine of the somewhat 

 long and slender abdomen 

 bear a pair of short lateral 

 spines. Of the five anal 

 spines, the ventral pair are 

 the longest, the lateral ones 

 shortest, and the dorsal 

 one blunt. Along the cen- 

 tral region of the dorsal 

 surface of the abdomen is 

 a broadish 

 vague line, 

 rather dar. 

 ker than 

 the ground 

 colour and 

 bearing a 

 central, 

 narrow, 

 slightly 

 lighter 



line. There are also on each segment a few tiny 

 black dots regularly placed. The processes at the 

 base of the fore-legs are close together, the anterior 

 one being about twice as long and but half as 

 broad as the posterior. 



Gordon Road, Kingston-on-Thames ; 

 January yd, 1895. 



B, processes at base of fore- 

 legs (x 15); a, anterior; 

 p, posterior. 



ILLUSTRATING SOUND-CURVES. 



By Joseph Goold. 



(Concluded from page 246.) 



A /T Y few explanatory remarks must be chiefly con- 

 *■**■ fined to what may be called ' ' musical curves" ; 

 that is, curves whose vibration ratios are identical, or 

 nearly so, with those of common musical intervals. 

 It should be remembered that the vibration figure 

 corresponding to a single musical sound is not a 

 curve but a straight line. This proposition will 

 not, however, bear reversing ; for the possible 

 varieties of motion in a straight line are infinite ; 

 and a combination of motion in two straight lines 



may represent or reproduce any plane figure 

 whatever — a fact most clearly demonstrated by 

 Professor Gray's "Tel-autograph." 



Let us now see what conditions determine the 

 forms of musical curves, especially those produced 

 by pendular motion. First consider the outlines or 

 initial forms only. 



The outline of a curve depends on the number 

 of vibrations in its constitution, and on their three 

 variable elements — ratio, phase and amplitude. 



