352 



SCIENCE 



[Vol. LV, No. 1422 



much of his time in doing routine work un- 

 connected with sciences, but he managed to 

 make a great number of interesting observa- 

 tions, which he hastened to commimicate to 

 Professor Poulton in letters, along with speci- 

 mens of most of the species referred to. He 

 hoped, on returning home, to work up his re- 

 sults and publish his more important discover- 

 ies, but he lost his life through a collision 

 at sea within a few hours of Liverpool. Pro- 

 fessor Poulton has edited his letters, adding 

 a brief memoir and numerous notes, together 

 with a series of contributions, from specialists, 

 describing many of the new or interesting 

 species found. 



The paper is so long, and its contents are 

 so varied, that it is impossible to give an 

 adequate summary. The principal section, 

 however, refers to the transformations and 

 habits of a number of species of Lycaenidae, 

 and brings out a number of new and curious 

 facts. It is a strange coincidence, that almost 

 simultaneously with Dr. Wheeler's publication 

 of the observation of beetles obtaining liquid 

 nourishment from coccids in South America, 

 Farquharson's account of similar habits in 

 Lyeasnid butterflies in Africa appears. The 

 butterfly concerned is Teratoneura isahellce, a 

 long account being given, showing that the 

 attending ants are driven away, apparently 

 by flapping the wings. Professor Poulton 

 suggests that an offensive odor is also pro- 

 duced. Later, two other related butterflies, of 

 distinct genera, were found to have the same 

 habits. Unfortunately the coccids were not 

 preserved, and we can only conjecture that 

 they were some species of Pseudococcus. 

 Both of the works reviewed were capable of 

 being completed only by the cooperation of 

 rather numerous specialists, entomologists and 

 botanists. It becomes increasingly evident that 

 much of the best work in bionomics must 

 necessarily be cooperative, no single individual, 

 however learned, being capable of dealing with 

 all of the species and problems involved. It 

 is pleasant to find, in the papers before us, 

 that the desired assistance was freely given 

 and is completely acknowledged. Only in this 

 spirit is it possible for men to woi'k harmo- 

 niously together, and any who fail to conform 



to proper standards should be made to feel 

 the disapproval of their colleagues. 



T. D. A. COCKERELL 



UNnrEBSiTT OF Colorado. 



SPECIAL ARTICLES 

 SEAUNG TUNGSTEN INTO PYREX 



The author has spent considerable time in 

 evolving a good method of sealing tungsten 

 wire into Pyres and fastening the copper lead- 

 wires to the tungsten. The method here des- 

 cribed is easily accomplished and the freedom 

 from breakage is certain. It is hoped that the 

 present detailed description may save others 

 sufficient time to justify its publication. An 

 elementary knowledge of glass-blowing is 

 assruned. 



The sealing-in glasses and the order of join- 

 ing are : 



tungsten— G70oH—G702P—Pyi-ex. 



The numbers are used by the Coming Glass 

 Company to designate these glasses. Some 

 glass-blowers prefer to omit G705H and seal 

 the tungsten directly to G702P. The G705H 

 is of lower melting point, may be used in the 

 gas-air flame and hence offers less chance to 

 oxidize the tung-sten. 



Clean the wire by sandpaper only or warm 

 in the flame, dip in a saturated aqueous solu- 

 tion of sodium or potassium nitrite (or nitrate) 

 and then polish with very flne sandpaper or 

 even the thumb nail. Draw small tubes of 

 each of the three kinds of glass having an in- 

 ternal diameter slightly larger than the dia- 

 meter of the wire. Cut a short length from 

 each and string them on the wire in the order 

 (above) in which they are to be sealed. The 

 flame should be applied first to the middle of 

 the G705H bead and the others in turn be 

 brought along the wire and melted to the pre- 

 ceding one. The wire with its glass coating. 

 Pig. 1-a, may then be sealed in in the usual 

 manner but joining Pyrex to Pyrex. In case 

 the tungsten wire is small and it is desired to 

 protect it from the flame the Pyrex enclosing 

 tube may be extended through the final seal. 

 Pig. 1-b, and the excess glass broken off after 

 the seal is accomplished. 



