ON THE CLIMATOLOGY OF AFRICA. 385 



to the Secretary of the Royal Meteorological Society, and may be freely 

 consulted by persons interested. 



Your Committee are under no illusion as to the merely conditional 

 value of many observations published by them. The index errors of the 

 instruments were unknown in many instances ; the hours for making 

 observations were injudicially chosen ; the observei's, owing to illness or 

 official duties, were frequently unable to fill up the registers, and there 

 was no one to take their place ; or, worse still, they had absolutely no 

 knowledge of the manner in which the instruments enti'usted to them 

 should be handled, and placed readings on record which, on the face of 

 them, are utterly absurd,^ and must unhesitatingly be rejected. 



Your Committee, on bringing their ten years' service to a close, 

 desire to direct the attention of the authorities called upon to organise the 

 meteorological service in British Protectorates or Crown Colonies to the 

 following points : — 



1. The instruments supplied should not only be verified before they 

 leave England, but should also be inspected periodically by a competent 

 official, who would pay particular attention to their exposure, inquire 

 into the competency of the persons charged with filling in the registers, 

 and eventually teach them how to observe. 



2. Inasmuch as all officials may occasionally be called upon to fill up 

 the registers, they should be instructed, before they leave England, in 

 handling and reading the usual meteorological instruments. An hour 

 spent at the office of the Meteorological Council, or with the Secretary of 

 the Royal Meteorological Society, would suffice for that purpose. 



3. It is of far greater importance to have a limited number of stations 

 well equipped, and the registers from which can be thoroughly trusted, 

 than a multiplicity of stations provided with defective instruments, care- 

 lessly or intermittingly attended to. 



4. Care should be taken that there should be no interruption in the 

 records kept at the principal stations owing to the illness or temporary 

 absence of the observer. Duly qualified rative assistants could be 

 obtained from the Meteorological Department of India. 



5. It is most desirable that the hours of observation recommended in 

 our ' Hints ' should be strictly adhered to, not for the sake of uniformity 

 only, but mainly because they yield a true mean of barometric pressure, 

 temperature, and humidity without making undue or unreasonable demands 

 upon the time of the observers. 



6. Unless local provision is made for the adequate publication of the 

 observations, the registers should be forwarded (through the Foreign or 

 the Colonial Office) to the Meteorological Council, or to the Secretary of 

 the Royal Meteorological Society, in order that abstracts may be prepared 

 and made generally accessible to meteorologists and others interested. 

 Still better would it be if an annual volume containing all these observa- 

 tions were to be published separately. 



' Not infrequently, as pointed out by us in publishing these observations, the 

 wet bulb and maximum thermometers give higher readings than the dry bulb and 

 minimum thermometers. Nay, some of these observers seem to be ignorant of the 

 decimal notation, for they enter 17'8 or 30"68 when there is no doubt that 17'08 and 

 30068 ought to have been entered. 



1901. C C 



