Mat 26, 1911] 



SCIENCE 



831 



It is to be noted that even these selected records 

 vary a hundred per cent. "Very notable variations 

 are found in the same mine or well, and often 

 much difference is found in adjacent records, es- 

 pecially those of artesian vpells. Some of these 

 are explainable, but the full meaning of other 

 variations is yet to be found.^° 



In conclusion, it may be stated that from 

 recent figures bearing on this subject, no gen- 

 eral law is observed in the increase of rock 

 temperature with depth, and in general the 

 increment of heat is lower and more variable 

 than indicated by the earlier observers. 



Thomas L. Watson 

 Brooks Museum, 

 University of Virginia 



/ 



' THE SCALES OF THE DIPNOAN FISHES 



I AM greatly indebted to. Dr. G. A. Boul- 

 enger for scales of the few living members of 

 the very interesting and remarkable subclass 

 knovm as Dipnoi or Dipneusti. The result of 

 their examination is quite surprising to me, 

 and has, I think, an important bearing on 

 their relationship with other fishes. 



Neoceratodus forsteri, from Queensland, 

 has very large oblong scales, one before me 

 being 55 mm. long and 34 mm. broad, 20 mm. 

 or less of the length being exposed in the 

 living fish. In general appearance, the scales 

 are not unlike those of Heterotis, except for 

 size. The broad nuclear field, far apicad of 

 the middle, is rugose; the circuli (fibrilte) 

 are all very fine, and both basal and apical are 

 longitudinal; the basal fibrillse are moniliform 

 (minutely tuberculate), and in the lateral 

 fields the whole surface is minutely rather 

 irregularly tuberculate. Thus in its longi- 

 tudinal fibrillse Neoceratodus agrees with 

 Amia and Alhula; in having the fibrillas 

 tuberculate or beaded it agrees with Albula 

 and the Osteoglossidse. At first sight it seems 

 that there are no radii in Neoceratodus, but 

 closer inspection shows a complete system of 

 fine radial reticulations, especially well devel- 

 oped in the lateral areas, where it accords per- 

 fectly with the network pattern of the Osteo- 

 glossids ! This exceedingly characteristic fea- 



" Op. cit., pp. 543-544. 



ture is now known, therefore, in the Dipnoans, 

 the Osteoglossids and the Mormyrids.' 



Having determined these facts, I turned 

 with eagerness to the material of Lepidosiren 

 and Protopterus. In these fishes the scales 

 are completely enclosed in the skin, but are, 

 nevertheless, quite large (fully 8 mm. diam- 

 eter in Protopterus), and shaped much as in 

 Osteoglossum. Both have a strong radial net- 

 work, while the circuli are reduced to innu- 

 merable fine tubercles or coarse granulations, 

 approaching the condition of the lateral areas 

 in Neoceratodus. Protopterus annectens from 

 Africa (Gambia) and Lepidosiren paradoxa 

 from Brazil have scales of entirely the same 

 type, but in the Protopterus the network is 

 more regular and more obviously similar to 

 that of the Osteoglossids. In both the fibril- 

 lar granulations tend to run in lines near the 

 margin, but this is rather more marked in 

 Lepidosiren; the indications are in each case 

 of longitudinal (not circular) fibrillse. The 

 general results may be thrown in the form of 

 a table, thus: 

 (_A) Basal fibrillee longitudinal. 



(a) Fibrillse moniliform or tuberculate. 



(1) With radial network . . . Dipneusti. 



(2) Without radial network. 



Albula and Dixonina. 

 (6) FibrillsB not tuberculate; no radial net- 

 work Amia calva, A. scutata. 



(B) Basal fibrillse circular (normal circuli) ; ra- 

 dial network present, 

 (o) Fibrillse tuberculate Osteoglossidse. 



(b) Fibrillffi not tuberculate . . . Mormyridse. 



It is also to be remarked that Gymnarchus 

 (Mormyridse), Heterotis (Osteoglossidas), 

 Lepidosiren and Protopterus all have larvse 

 with external gills. 



Dr. Boulenger has very kindly sent me the 

 scales of the Osteoglossids Scleropages for- 

 mosus from Borneo, Scleropages leichardti 

 from Queensland and Osteoglossum bicirro- 

 sum from Cadajos, Brazil. They are prac- 

 tically circular (S. leichardti rather broader), 

 and all have exactly the same structure, not- 

 withstanding the wide geographical separa- 

 tion. The scales of Heterotis niloticus diiier 



^ For the last, see Smiths. Misc. CoU., Vol. 56, 

 No. 3, p. 2. 



