Januaky 17, 1908J 



SCIENCE 



105 



water with mercury thermometers and had 

 failed to discover any real variations from the 

 freezing point, but the Callendar platinum 

 resistance thermometer makes it possible to 

 determine the variation of the temperature 

 easily to a thousandth of a degree centi- 

 grade, and Dr. Barnes's familiarity with this 

 instrument enabled him to make extremely 

 accurate observations showing the condition 

 existing when frazil and anchor-ice are 

 formed. 



The first few chapters of the book are de- 

 voted to a compilation of the laws of the 

 transfer of heat, the various methods of de- 

 termining the different constants of ice and 

 water, and the general structure of ice. In 

 referring to the various workers in these fields, 

 Dr. Barnes gives their names and the year of 

 their work hut does not give references to 

 their original publications, which would be 

 very useful for students of the subject. The 

 platinum resistance thermometer is then de- 

 scribed and the degree of accuracy obtainable 

 with it is shown to be in the neighborhood of 

 1/10,000 of a degree centigrade, under good 

 laboratory conditions; these were not obtain- 

 able at the river bank where Dr. Barnes made 

 his observations, but he considers that the 

 water temperatures which he determined are 

 accurate to a thousandth of a degree. The 

 rest of the book is devoted almost entirely to 

 the study of anchor ice and frazil, and their 

 occurrence in the St. Lawrence River. 



Many observations and opinions regarding 

 the formation of anchor ice are cited and Dr. 

 Barnes wisely makes a long extract from two 

 papers by the Rev. Dr. Farquharson, in the 

 Philosophical Transactions for 1835 and 1841. 

 His observations were made on the rivers Don 

 and Leochal in Aberdeenshire, and they were 

 so well made that one is entirely convinced 

 that he is right in ascribing the formation of 

 anchor-ice to the cooling of the bottom of the 

 streams by radiation. The absorption of 

 water for radiation of long wave-lengths is 

 not well known but Dr. Barnes made a pre- 

 liminary experiment to show that bodies at 

 low temperatures can radiate through water 

 and glass; a coil of platinum wire covered 

 with a layer of water between glass, was ex- 



posed to radiation on a clear night; the coil 

 became cooler by radiation. 



Careful observations made near Montreal 

 show that anchor-ice is not formed except on 

 cold clear nights and in situations permitting 

 free radiations into space. Anything inter- 

 fering with this radiation, such as a bridge or 

 a clouded sky, will entirely prevent the forma- 

 tion of anchor-ice, and a very short period of 

 bright sunshine is sufficient to loosen the ice 

 from the bottom and cause it to rise, some- 

 times in great quantities, to the surface. 



Frazil, on the other hand, is formed at the 

 surface, where the water is so disturbed by 

 currents or strong wind as to prevent the 

 formation of a surface sheet. All that is 

 necessary for the formation of this kind of ice 

 is very low temperature and rough water. 

 The water becomes undercooled to a few 

 thousandths of a degree and small crystals of 

 ice are formed and carried down with the cur- 

 rent. Bright sunshine is sufficient to prevent 

 the formation of frazil; and the experiment of 

 Dr. Barnes shows how quickly the under- 

 cooled water is brought up to the freezing 

 point when the sun shines upon it. His ex- 

 periments were carried out in a small hut on 

 the shore of the river. It is necessary in 

 these experiments to have as a basis for com- 

 parison a mixture of snow and water exactly 

 at the freezing temperature; some difficulty is 

 found in obtaining this to the degree of ac- 

 curacy required. Dr. Barnes found that when 

 the water was freezing the temperature of the 

 mixture was slightly lower than the freezing 

 point; on the other hand, when the ice was 

 melting the temperature was slightly higher, 

 and he found that the variations from the 

 freezing point depended upon the relative pro- 

 portions of ice and water in the mixtures. 

 This led him to the suggestion that " a freez- 

 ing point mixture must be defined as an inti- 

 mate and equal mixture of ice and water, 

 neither gaining nor losing heat." 



The final chapter in the book tells of the 

 difficulties attending the use of water-power 

 machinery at Montreal on account of the 

 formation of frazil ice, and the way these have 

 been met. One very simple method, which has 

 served pretty well, is to take the water from 



