THE AMERICAN BOTANIST. 67 



plants. The round bodies in the axils of the Hly leaves above 

 ground have the same relative position as the bulblets below 

 ground and in fact are really a second form of bulb — or trans- 

 formed branch or bud just as you please. In certain lilies 

 we have even better evidence that these objects are transformed 

 branches, for in these the bulbs are not formed in the axils of 

 the leaves but instead a branch occurs there which after grow- 

 ing some distance forms a bud or bulb at the tip. The adder's- 

 tongue often sends out half a dozen such branches from a 

 single bulb and thus one bulb becomes several by the end of 

 a year. If these branches should be slender and their tips be 

 inordinately stocked with food, we should have a tuber, such 

 as other plants often produce. Here again we see the small 

 differences that separate one form from another. The house- 

 leek, which sends out a branch above ground producing a 

 tuft of leaves and ultimately a new plant at the tip, comes 

 very near to producing a bulb above ground or, since it is 

 thickened with food stores, we might, perhaps, call it a tuber. 

 In certain onions, also, may be seen another bulb above ground 

 where such bulbs may occur among the flowers or the whole 

 flower-cluster may be turned to bulbs. It is well known that 

 the flower, itself, is a greatly modified branch, and the bulblets 

 thus produced go to prove it. 



Plants and Freezing. — The living parts of plants con- 

 sist of very minute box-like structures called cells lined with a 

 clear watery substance like the white of an egg, called proto- 

 plasm. This protoplasm is the real living part of the plant 

 and in some species is very susceptible to cold and dies as soon 

 as the temperature approaches the freezing point. Many 

 tropical plants belong to this class. In other plants, notably 

 the perennials of cold regions, mere cold, unless extreme, does 

 not kill the protoplasm, though sudden freezing and thawing 

 may do so. The reason for this is found in the fact that in 



