HOW A NATURALIST IS TRAINED. 633 



and its relative importance determined, before conclusions can be 

 drawn, and even tben it is frequently necessary to estimate the rela- 

 tive probability of two or more alternatives, tbus giving an exercise 

 to the powers of ratiocination which is utterly lacking in the remorse- 

 less logic of the mathematical theorem. 



When this general foundation of facts and methods is obtained, 

 more special studies are taken up ; and since embryological research 

 includes most of the processes involved, we will suppose that the stu- 

 dent is next introduced to this fascinating field which is now so assid- 

 uously cultivated by scholars all over the world. As a rule, it may be 

 stated that animals living on the land or in fresh water differ consider- 

 ably in their mode of development from their near relatives in the sea. 

 Many ancestral features which are retained in marine forms have be- 

 come eliminated in the others, and hence the study of the growth of 

 salt-water forms from the egg to the adult throws far more light on 

 the relationships and ancestry of the different groups than does that 

 of the terrestrial and fluviatile species. Again, the sea affords a wealth 

 of life far beyond that of the land and fresh water, a wealth to be esti- 

 mated not only in number of individuals but of species as well. Whole 

 groups of animals are solely marine, while others are represented 

 on the earth or in rivers and ponds by a few small and insignificant 

 forms. 



For this reason the student of embryology betakes himself every 

 year to the shore, so that, while being recuperated by the sea-breeze, 

 he may continue his studies and add to the total of human knowledge. 

 Marine laboratories for this purpose are scattered the whole length of 

 our coast, from North Carolina to Eastport, some being mere tempo- 

 rary affairs, others permanent stations. In 1885 public or private 

 laboratories existed at Beaufort, Newport, Nantucket, Wood's Holl, 

 Salem, Annisquam, Mount Desert, and Eastport. The absolutely 

 necessary furnishings of such a laboratory are extremely few, but to 

 them one may add as far as purse and inclination admit. There must be 

 tables, chairs, and glass dishes, while each student must have "a micro- 

 scope and accessories. Then come a boat (a row-boat is sufficient for 

 all ordinary work) and apparatus for collecting. Finally, a small stock 

 of chemicals and apparatus for microscopical work complete the list of 

 necessities. 



Possibly the most common way for obtaining material for embryo- 

 logical study is by use of the skimming-net. This consists of a brass 

 ring about a foot in diameter, to which is attached a net of fine gauze 

 and cords for dragging it behind a boat. The whole operation of 

 skimming is very simple. Two persons are required, one to row the 

 boat, the other to attend the net. The latter allows the net to trail 

 behind, keeping the cords so that part of the mouth is above and part 

 below the surface of the water, so that as much as possible of the sur- 

 face-water will be strained through the gauze. At intervals the net is 



