i 4 ROUND THE YEAR 



graph snow-crystals, and this has been done with 

 great success by Neuhauss and Nordenskiold. 1 It is 

 obvious that photographs record many particulars 

 which cannot be accurately recorded by drawings, 

 made from objects so fugitive as snow-flakes. They 

 also preserve many departures from symmetry which 

 have been neglected in the drawings made by hand. 



The large flakes, with a diameter of half or three- 

 quarters of an inch, are not themselves snow-crystals, 

 but aggregations of such, sometimes very loosely 

 attached. A diameter of a fifth of an inch is very 

 large for a snow-crystal, and the average diameter of 

 snow-stars, the commonest form, is only about one- 

 tenth of an inch (2.35 mm.) Other snow-crystals are 

 yet smaller. Keen sight or the help of a lens is 

 therefore necessary to make out the exact shape of a 

 snow-crystal, and all the finer details require the 

 microscope. 



When we examine a snow-crystal carefully, we soon 

 learn one fact respecting it, viz., that it is six-rayed. 

 The crystal forms along three lines or axes lying in 

 one plane, which cross each other at equal angles. 

 Six lines proceeding from a common point at equal 

 angles will, of course, be 60 apart. There is a fourth 

 axis to the snow-crystal, which we do not recognise at 

 first. This represents the thickness of the crystal, 

 and takes a direction at right angles to the plane 

 in which the other (lateral] axes lie, passing through 



1 Neuhauss's figures are reproduced in Hellmann's Schnee- 

 krystalle, a useful and interesting book, which has furnished the 

 materials for this historical sketch. A number of Nordenskiold's 

 figures are reproduced in facsimile in Nature, October 19, 1893. 



