36 



SCIENCE 



[N. S. Vol. XXIX. No. 731 



Edward Bovalius of the Essequibo Explora- 

 tion Company, 



Progress above Cangaruma was retarded 

 slightly by fever. A portage at Amatuk and 

 another at Waratuk with two days paddling 

 brought the expedition to Turkeit at the base 

 of a series of Cataracts at the head of which 

 is Kaieteur Eall, the show place of British 

 Guiana. The Potaro here leaps 741 feet from 

 the plateau to the Potaro Gorge below. 



A portage of 2J hours carries one from 

 Tukeit to the Savannah above the falls. Mr. 

 Schideler returned from this point to collect 

 near the coast, while Professor Eigenmann 

 continued with thirteen Indians to Holmia 

 and Arnataima, the first cataract above 

 Holmia. 



Thorough collecting was undertaken in the 

 Potaro above and below the Kaietue with the 

 use of hooks, seines, dynamite and most 

 effectively with Haiara, the poison used by 

 the natives in their fishing. The expedition 

 was thoroughly satisfactory although few 

 photographs of fishes could be taken on ac- 

 count of the labor necessary to secure suitable 

 specimens. 



It is hoped that it will be made possible to 

 continue the exploration of the Guiana 

 Plateau which sends rivers over heavy falls to 

 the Essequibo, Orinoco and Amazon from 

 around Eoraima, possibly the oldest land area 

 of South America. 



The bit of most satisfactory discovery is 

 that Gasteropelecus, an aberrant characin, 

 flies. This fish has the most amazing struc- 

 ture for a characin. It possesses huge 

 pectorals, a tremendous " sternum " and 

 pectoral muscles to correspond. It was fre- 

 quently seen to dart from in front of the boat, 

 float its pectorals while part of its tail and 

 sternum remained in the water and then in 

 the last five or ten feet of its 45-foot flight 

 clear, the water. As long as part of the fish 

 remains in the water the pectorals touch the 

 water with each stroke. Not the least in- 

 teresting fact is that their line of evolution 

 from generalized Characins is indicated by the 

 still-existing genera like Chalceus and Pseudo- 

 corynosoma. 



SPECIAL ARTICLES 

 SPECTRUM OP COMET MOEEHOUSE 



The spectrum of this comet has been under 

 observation here since October 28, with the 

 use of a Zeiss photographic doublet of 145 mm. 

 aperture and 81 cm. focal length, made of 

 " ultra-violet " glass, over which was placed a 

 15° objective-prism of the same glass and 

 aperture. With an exposure of fifteen 

 minutes the head of the comet then gave a 

 sufficiently strong impression, showing a row 

 of seven knots. With longer exposure the tail 

 could be well traced from some knots until it 

 ran off the plate, at a distance of 3°. 



The absence of a continuous spectrum was 

 striking, and on no plate thus far obtained haa 

 it been certainly visible. This indicates that 

 during this period the reflected light has been 

 exceedingly weak in both head and tail rela- 

 tively to the intrinsic light due to the carbon 

 and cyanogen bands. 



A very small quartz spectrograph was also 

 attached to the same telescope, and plates hav- 

 ing the advantage of a comparison spectrum 

 were obtained on four nights. Twenty-one 

 satisfactory plates were obtained with the ob- 

 jective-prism on eleven dates. All the plates 

 were made by Parkhurst, with assistance, when 

 necessary, from Frost. 



The measurements of wave-length are quite 

 uncertain on account of the very small scale 

 of the spectrum with either apparatus — only 

 2.5 mm. from H p (A 4861) to E 9 (A 3798) 

 with quartz spectrograph (3.0 mm. with ob- 

 jective-prism) — but we have been surprised at 

 the accordance of our measures on different 

 plates. It is difficult to mal^e settings on the 

 edges of the cometary bands on account of 

 their diffuseness, so that it is often necessary 

 to be content with settings on the centers of 

 the loiots. 



We regard the identification as certain for 

 the third and fourth carbon bands (edges at 

 A 5165 and A 4737) and the first, third and 

 fourth cyanogen bands (A 4601, 3883, 3590). 

 These carbon bands are two of the three bands 

 characteristic of cometary spectra, and often, 

 perhaps generally, ascribed to a hydrocarbon. 

 The other one, at A 5635, did not affect our 



