54 



KNOWLEDGE. 



IMabch, 1903. 



trausportiug his fellows, which number some thousands. 

 To conveuieutly bear this burden, the body of this second 

 " Atlas " is drawn out into a long stalk-like stem, and 

 around this the various units of the colony are grouped ; 

 whilst the upper end of this stalk is dilated to form a 

 large and somewhat pear-shaped bag which, filled with air, 

 forms the float or life-buoy. Immediately below the float 

 come what we may call the motor members of the colony. 

 These are bell-shaped, and by alternately opening and 

 closing, draw in and expel the water from their cavities 

 and thus propel the whole colony through the water. These 

 motorists, like the central member which support* the whole, 

 take no food, neither jiart in the production of the young ; 

 they have relegated these duties to other and specially 

 c[ualified members. The capture and preparation of the 

 food is undertaken by peculiarly modified members found 

 beneath the swimming members. Scattered among the 

 motorists are found numberless small oval individuals, 

 each with a long slender thread. These threads are 

 extremely sensitive, and by their means the members 

 bearing them are enabled to uudertake the duties of feeling 

 and exploring for the colony. Lastly, there are the 



Fig. 2. — Phiisaha, the " Portu^iese Man-o'-War," 

 floating on the surface of the water. 



individuals whose sole duty is to provide for the perpetua- 

 tion of the species. And all these, as we have before 

 remarked, are borne by a single member designed by the 

 Fates, for the sake of the colony, to become a wind-bag. 



The wonderful and dread " Portuguese man-o'-war " of 

 the blue waters of the Mediterranean also proves, when 

 examined, to be really a floating colony similar to that of 

 Physophora. But the float or life-buoy in this case is 



much larger, taking the form of a pear-shaped bladder 

 provided with a many-chambered crest or comb. If its 

 eluties are of the menial order, it has at least ((iinpensation 

 therefor, inasmuch as it shimmers in the sunlight with 

 the most gorgeous c<>k)uring, looking like molten silver 

 tinted with light blue, violet, and purple, whilst the 

 small thickenings in the comb are made conspicuous by 

 vivid carmine. The long feelers may measure as much as 

 50 feet in length, an<l are vested with a stinging power 

 which makes this creature more dreaded by the native 

 than the shark. 



Finally, we may mention certain Mollusca which have 

 adopted the air-float for the purpose of supporting the 

 eggs. Thus the beautiful oceanic violet sea-snail, lanthiua, 

 throws out a long raft containing the eggs, which are 

 carried along the surface of the water, where they may 

 catch the sun and air. It is to be noted, however, that this 

 float difl'ers materially from those of all the others which 

 we have examined, inasmuch as it is made by imprisoning 

 bubbles of air in a gelatinous secretion, and not by the 

 inflation of a pouch or wind-bag. 



GIANT L^ND-TORTOISES. 



By R. Lydekker. 



In the long-past days when the plains of India were the 

 home of the mighty sivatherium and of still more gigantic 

 elephants and mastodons, while its rivers were tenanted 

 bv hippopotamuses and huge long-snouted, gavial-like 

 crocodiles, that country was likewise inhabited by the 

 most gigantic land-tortoise of which we at present have 

 any knowledge. When fragments of its fossilised shell 

 and more or less nearly complete specimens of its limb- 

 bone came under the notice of its original descrilsers, it 

 was thought, indeed, that they indicated a creature of 

 truly colossal proportions, the length of the shell in a 

 straight line being estimated at no less than twelve feet 

 three inches. In a restoration of the shell made under the 

 superintendence of the discoverers of the species, and still 

 exhibited in the geological department of the Natural 

 History Museum, the length was reduced to a little over 

 eight feet. But even these reduced dimensions appear to 

 be considerably in excess of the reality, and it is probable 

 that the maximum length did not much exceed six feet. 

 A shell of this size vastly exceeds, however, that of any 

 modern laud-tortoise, so that the Siwalik tortoise, or 

 Testudo atlas, as it is scientifically called, is fully entitled 

 to rank as the real giant of its kind. 



But the Siwalik tortoise was by no means the only giant 

 species inhabiting India during the Pliocene epoch, as 

 remains of other, although smaller, forms have been 

 discovered in the same deposits. The nearest living ally 

 of the Siwalik species appears to be Testudo emys, of the 

 countries east of the Bay of Bengal, in which the shell 

 does not much exceed a foot in length. Both kinds have the 

 front end of the lower shell produced and notched, 

 although the production and notching are much more 

 pronounced in the extinct form. Both also have the horny 

 shield immediately above the tail double, instead of (as is 

 usually the case) single ; and in both the skin of the legs 

 contained embedded nodules of bone. 



The Pliocene deposits of the south of France have also 

 yielded remains of a giant land-tortoise (T. perpiniana), 

 with a shell abovit four feet in length, and likewise furnished 

 with bony nodules in the skin of the limbs. And from the 

 caves of Malta have been obtained bones of yet another 

 very large species (T. robusta), apparently allied to the 

 recently extinct T. inepta of Mauritius. 



Going further afield, we find evidence of the existence 



