78 



SCIENCE- G OS SIP. 



cement over the inside of the tube was continued. 

 It was clear that the spinning was done wholly 

 with the first and second peraeopods, the tips of 

 which were touched from point to point over the 

 inside of the skeleton tube in a way that recalled 

 strongly the movements of the hands in playing 

 upon a piano. The cement adhered at once at the 

 points touched and spun out between them in 

 nnif orm delicate threads, which appeared to harden 

 quickly, and did not seem, even at first, to adhere 

 to the animal itself. In a case in which the entire 

 construction of the tube was watched, the Micro- 

 deutopus very nearly or quite completed the work 

 in a little more than half an hour. In the Amphi- 

 thoe, Smith adds, the process of constructing the 

 tube was very similar, though less cement and 

 more foreign material entered into the structure ( 12 ). 

 Giles, who watched Cera/pus calamicola under the 



microscope, one day surprised an individual in the 

 act of repairing the fibrous lining of its reed-tube ; 

 the animal had completely withdrawn into the 

 tube, and was keeping the latter slowly but con- 

 tinuously revolving. The tube was transparent 

 enough for the observer to see that the Cera/pus 

 remained stationary while the tube revolved ; but 

 he was unable, unfortunately, to make out the 

 exact manner in which the fibre was being de- 

 posited. It may be noted, finally, that this 

 naturalist supposes certain teeth in the animal's 

 second gnathopods to be well suited for cutting 

 the secreted thread, or for trimming the piece of 

 reed used for the foundation of the tube ; and he 

 further suggests that the distal segment of the 

 third peraeopods is admirably adapted for guiding 

 the thread. 



(To be continued.') 



DESMIDS. 

 By Dr. G. H. Bryan, F.R.S. 



{Concluded from Vol. VI. page 360.) 



A CCORDING to the classification in Dr. M. C. 

 -*--*- Cooke's " British Desmids " there are twenty- 

 one genera represented in Great Britain. In the 

 following columns it is proposed to give a, super- 

 ficial account of the differences between the genera, 

 rather than a detailed description of their botanical 

 features, as the above-named treatise supplies all 

 that is wanted for a fuller study of the desmids. 

 The genera are divided as follows : — 

 Section A. Leiosporae. Zygospores usually 

 smooth. 



Sub-section a. Individuals more or less closely 

 united in threads or filaments. Genera : 

 Gonatozygon. Sphaerozosma, Onychonema, 

 Hyalotheca, Bambusina, Besmidium. 

 Sub-section b. Cells free, not united in a fila- 

 ment. Genera: Docidium, Closterium, 

 Penium, Cylind/roeystis, Hesotaenium, Tetme- 

 morns, Sp\rotaenia. 

 Section B. Cosmosporae. Zygospores normally 

 warted, spinulose or ornate. Genera : Micras- 

 terias, JSuastrum, Cosmocladium, Cosmo/Hum, 

 Calocylindrus, Xa nth idi u m , A rtli rodesm us. 

 Staurastrum. 

 Commencing with the filamentous forms, the 

 genus Gonatozygon contains three species, all of 

 them apparently local, and considerably resembling 

 confervoid algae with their cylindrical filaments 

 formed of elongated cells. Some of the species of 

 'Sphaerozosma and the two British species of 

 Hyalotheca are enclosed in a thick gelatinous 

 sheath which is somewhat difficult to see in un- 



(12) Smith, 1882, I.e. pp. 274, 275 ; " Mature," xxii. (1880), 

 pp. 594, 595. 



stained preparations, and is practically invisible in 

 glycerine. I have been recommended to use 

 Bismarck brown as a stain, but have not yet had 

 an opportunity of trying that pigment. The 

 envelope may be twice or thrice or more times the 

 diameter of the central filament of cells. While 

 the cells of Hyalotheca are cylindrical, those of 

 Sphaerozosma are aptly described by the late Rev. 

 J. G. Wood, in his " Common Objects," as "looking 

 much like a row of stomata set chain-wise together." 

 In the single species of Onychonema recorded 

 from Strensall Common, near York, the alternating 

 and overlapping horns are characteristic. Both 

 Hyalotheca and Sphaerozosma are represented in 

 this neighbourhood, the latter genus occurring 

 in the Sphagnum washings from the Xant Ffrancon 

 valley. The widely distributed sole representa- 

 tive of Bambusina has curiously barrel-shaped 

 cells. In Besmidium we have triangular cells placed 

 one above the other, forming a filament in the 

 shape of a twisted prism. To get an idea of this 

 arrangement it is only necessary to take a pile of 

 books and twist them slightly from the top down- 

 wards. This twist gives the appearance of a spiral 

 band formed by the angles of the cells running 

 diagonally across the filament. 



Passing on to the " free " desmids, we find in 

 Docidium an elongated straight cell divided into 

 two segments by a constriction in the centre, and 

 having the same peculiar bodies at the tips of the 

 cells that are found in the horns of Closterium. 

 The genus Closterium contains about forty British 

 species and varieties, two of which are figured in 

 Scienxe-Gossip for April last (page 325). The 



