342 



SCIENCE PROGRESS 



If we have a tetraster of four consecutive pole-centroids l at 

 the apices of a square, and introduce a permeable sphere at its 

 centre, we obtain a beautiful pentaster with five centroids and 

 eight spindles (fig. 9, a). 



Another polyaster might be obtained if we could place four 

 like poles at the vertices of a tetrahedron, and a permeable 

 sphere at its centre ; the actual plane model is shown below 

 (fig. 9, b). 



This is perhaps the best place to consider the " false 



Fig. 8. — Triaster (dust on paper) formed in a circular field on three interposed centroids. 

 The current passed up through three meridional bundles and down through the axial 

 bundle shown ; and the paper lay in the equator of an approximate sphere. 



spindles " of Rhumbler, Leduc, and Gallardo (in his second 

 interpretation). Leduc has made models with diffusion currents 



1 The odd-figures, the tri- and pentaster, are the most important from a 

 theoretical point of view ; for Rhumbler, not recognising the action of an 

 interposed centroid in determining the insertion of spindle-ends, had declared that 

 magnetic fields — and indeed this reasoning applies to all dual forces — could not 

 possess an odd number of centres and of spindles. " Magnetic lines of force can 

 never form three spindles between three consecutive poles . . . three adjacent 

 poles, even when there are more than three poles in the field, can never be joined 

 up by three spindles " (" Mechanische Erklarung der Ahnlichkeit zwischen 

 Magnetischen Kraftliniens Systemen und Zelltheilungsfiguren," Arch, Entwick- 

 lungsmeck, vol. xvi., 1903, p. 482). 



