Vol. VIII. Xo. I si. 



THE AQRICULTUEAL NEWP. 



105 



Supports for Yam Vines. 



The practice ot' ' staking yams,' i.e., <it' providing 

 stakes or other supports on which van) \ines may 

 climb, is commonly followed in islands where the 

 existence of abundant woodlands rcndeis the stakes 

 easily obtainable. It is evitlent that 'staking' is 

 especially instrumental in increasing the yield on 

 retentive soils, and where the rainfall is heavv'. WJien 

 supports are jirovided for the vines, closer planting can 

 be adopted, and the land economizi:<l in this way. An 

 experiment bearing on this point was caiiiiil out last 

 year at the 8l. Lucia Agricultural School. Two kinds 

 of yanus — 'Lisbon' and 'Bottle-neck Lisbon' — were 

 grown both staked and unstaked. The 'Lisbon' yam 

 yielded at the lato of o'2 tors poi' acre when unstaked, 

 and 7 tons when staked. The st.died plot of 'Bottle- 

 neck Lisbon ' gave at the late of 43 tons per acre, and 

 the unstaked plot only 24 tons. The deficient yield 

 from the unstaked plants is due to tlie fact that, when 

 planted at ordinary distances, tiie yam vines are 

 produced in such profusion that it is impo.ssible for 

 light and air to have proper access, and hence full 

 development cmnot take place. 



Good Varieties of Yams. 



Experiments with a number of varieties of V'ams 

 have been cai'ried on annually for many years past 

 both at Antigua and St. Kitt's, and. as a result, several 

 kinds have now definitely proved themselves to possess 

 a crop-yielding capacity above the average. Among 

 these are 'Light Red,' 'Sealed Top,' 'Bottle-neck 

 Lisbon,' ' Bugle Horn,' and ' Cush Gush.' 



At Antigua the yields have been as follows, on the 

 average of the past five years : 'Light Red,' 10,7.S0 lb. 

 per acre ; ' Sealed Top,' 7,700 lb. ; ' Bottle-neck Lisbon,' 

 7,32() tb. : 'Cush Cush,' 6,314 It)., and ' Bugle Horn,' 

 (3,182 Ih. It sho)dd be mentioned that these results 

 are in every case, the average from two plots, one of 

 which was mantired, while the other was unmanured. 

 (!)n the whole, however, there was little increase in the 

 crop consequent upon n)anuring. 



The order of merit of the different varieties has 

 been very similar at St. Kitt's to that at Antigua, but 

 higher v'ields have been obtained in many cases. Thus, 

 ' Light Red ' gave a retui-n at the rate of 14,G2G lb. 

 of yatns ]jer acre, ' Sealed Top,' I3,(S49 ft. per acre, 

 and 'Cush Cush,' cS,781 ft. per acre. 



Cotton Seed Selection Experiment. 



An e.xpoiment in seed selection, the results of 

 which should be of interest to cotton growers, is 

 desciibed in the last annual report of the Curator of 

 the jMontserrat Botanic Station. 



It is known that clean, black cotton seeds invariably 

 bear a lint which is poor and wast}' in character. 

 Fiirt.h(!r, it has been observed that the character of the 

 seed in the case of any given Sea Island cotton plant 

 is uniform, i.e., clean, black seeds and ' fuzzy ' seeds 

 are not found on the same j)lant. It was desired to 

 ascertain, therefore, whether the clean black, and the 

 ' fuzzy ' seeds repioduced thetnselv^s true to type or 



not. Two jilots were prepared, and one was sown 

 with clean, black cotton seeds, while the second was 

 i)Ianted uith seeds liavinsr an average amount of f))zz.' 

 The results were that, on the first plot, out <f 170 

 cotton pl.tnts. 120 bore seeds which were of a clean black 

 character, and similar to those sown ; fort3'-fi ve plants had 

 seeds with a slight amount of fuzz, and five plants with 

 a fair amount of fuzz. On the second plot (sown with 

 f)izzy seeds), out of 150 phmts, 134 boie seeds similar 

 to those f)lanted, in that they had an average atnount 

 of fuzz The remaining si.xteen plants yielded seed 

 having little or no fuzz. 



This experiment shows the advisability of re)nov- 

 ing all clean, black cotton seeds before planting, as in 

 this way only a small percentage of such seed maybe 

 expected froto the resulting plants. 



' Black Blight ' at Grenada. 



•Black blight' is probabl}' in more general 

 evidence on numbers of different trees and plants at 

 Grenada than in any other of the West Indian islands. 

 The so-called 'blight' itself, it will be remembered, 

 is rather a symptom than a cause of damage to plants on 

 which it is noticed, since the fungus is found in associa- 

 tion with scale insects, on the excreta or 'honey-dew' of 

 which it feeds, doing no direct harm to the trees. 



The Agiicultural Superintendent at Grenada 

 reports that )nuch has lately been done by individuals 

 in the island in the way of clearing their estates of 

 blight. Useless trees attacked with it have been 

 destroyed, others have been lopped and white-washe<l, 

 while in a few instances some spraying has been done. 

 The great difficulty experienced — as in all matters of 

 this sort — is to command any co-operation in attacking 

 the pest, and w-ork done on certain estates is often 

 spoiled by neighbours who will do nothing. 



Cows and Their Milk Yield. 



Experimental work carried out in many parts of 

 the world have now thorotighly established the fact 

 that the milk-yielding capacity of a cow is practically 

 pre-determined by the breed and strain of the animal, 

 and that while the milk yield may vary in quantity or 

 q)iality within narrow limits as the result of different 

 systems of fee<ling, yet food is a factor of secondary 

 importance in this connexion. The most plentiful anJ 

 well-balanced rations will not enable a Holstein cow to 

 give the rich milk of a Jersey, and, on the other hand, 

 if Guernse}' or Jersey cows are fed on a poor diet, 

 althotigh their inilk ma}' decrease in quantity, yet th^e 

 amount given is still rich in butter-fat. Food has 

 a greater infiiience upon the tpiantit}' than the quality 

 of milk yielded. Every cow has a certain maxim)im 

 capacity which, as already inelitioned, is dependent on 

 its strain and natural character. Under proper condi- 

 tions, and with stiitable feeding, this maximum is 

 reached, and beyond that, food has no influence. This 

 indicates the in)portanc'. yf good bre<?d. and strain ir» 

 niilkin£r co\\.«. 



