258 



KNOWLEDGE 



[November 1, 1890. 



various foi-ms of which are given, and then the frank 

 statement that " no related words are known, even in 

 Teutonic, except its own derivatives : see Breadth, 

 Brkde." 



After such a confession by this high authority, I shall 

 surely ofiend no prejudices by venturing to suggest some 

 " related words " which appear to have been overlooked. 



By the simplest form of metastasis [cf. hrydde = bird'] 

 we get to the word Ilaard, a broad extended surface of 

 wood, a word for which the Dictionary offers nothing 

 better than tentative suggestions as to its correlations. 

 The Dutch word hard is used for a plate or trencher, that 

 being an extended surface ; and the old English word Bird 

 for " a board, or tablet," gives another congener of Broad 

 or Board. But what about the word Bread ? This word 

 is left to surmises by the Dictionary, because it did not 

 occur to the writer that hrati derived its name from the 

 shape of the broad flat cakes which were the first repre- 

 sentatives of the loaf. The failure to recognize this 

 alliance is the more remarkable when we remember that 

 the word bread survived in its primitive sense (when 

 applied to ships' biscuits) do^^Ti to the 18th century. A 

 similar word is Brad, which the Dictionary itself defines 

 as " a thin rlaltish nail '' ; its liro((dened shape being the 

 cause of its name. 



The common interchange of I and r accounts for the 

 old word Blad, " a firm fat blow," and this brings us to 

 Blade, a word always expressive of that which is broad, 

 whether it be a blade of gi-ass, the blade-bone, or the 

 blade of a knife, sword, or paddle. This is clearly seen in 

 the southern Scotch use of the word blade for the broad 

 outer leaves of cabbage, lettuce, &c. ; and in the German 

 restriction ol blatt to a leaf, while Inub (i.e. leaf) is applied 

 to foliage in general. And do not the words liate. Platter, 

 a Plot of ground, and a gra.ss-Plat, give us other variants 

 of the interchangeable letters ji-l-t, b-l-t, b-l-d, b-r-d, in 

 the sense of extended surface '? 



The well-known change of p to f brings the word Flat 

 itself into the series ; for is not every flat surface a broad 

 or extended one ? Surely no philologist of the narrowest 

 school will deny that the Greek TrAan's is basically con- 

 nected with the word }date, and that both are allied'to the 

 Sanskrit jirath, " to extend." The Sanskrit word for 

 broad is pritlm, in which we find the labial, semi-vowel, 

 and dental, just as in the European languages ; and 

 pritliwi is the name of the earth because of its extended 

 surface. 



The fi-amework of all the words here adduced is — 



It is not pretended that these words are derived from 

 each other, but that they are most certainly basically con- 

 nected together, and may fairly be used to modify the 

 statement that " no related words [of Broad] are known, 

 even in Teutonic, except its own derivatives." 



Frederic Pincott. 



WET DATS IX SUMJIER. 



To the Editor of Kxowxedge. 



Sir, — Looking through the records of rainfall at Green- 



wich since 1841, I have met with a fact which I do not 



remember to have seen pointed out before, and which 



seems to deserve the attention of meteorologists, farmers, 

 and others. 



It is, that the character of June, as regards rain, 

 is to some extent a key to that of the rest of the season 

 (meaning by " season " the four months .Tune to Septem- 

 ber). More often than not, if .June is wet, the three other 

 months (viewed as a whole) are wet ; if dry. dry. We 

 may, if we like, ft-ame a statement which has wider, and 

 nearly universal, application, thus : — 



11 here ue jind a iret June ire jind a iret season : where a 

 dry .June, a dry season. 



Here we include those cases in which the excess or 

 defect of Jime makes the season wet or dry, though the 

 remaining three months may be average, or dry (in the 

 case of excess), or wet (in the case of defect). 



I have measured wetness by the' number of wet days, 

 meaning by a wet day one on which at least one-tenth of 

 an inch has fallen. A wet month, season, or other period, 

 will be imderstood to mean one with more than the 

 average number of wet days. 



If anyone will take the trouble to make out a curve of 

 the number of wet days in June, and another of the wet 

 days in the whole season, or in June-September, he will 

 find a remarkable correspondence. 



Those forty-nine seasons may be classified in the fol- 

 lowing way, which brings out very well the matter to be 

 explained : — 



Reckoning 



49 



40 



I have added a column in which all days of recorded 

 rain are considered, and the results seem stDl better than 

 in the first column. 



Thus, excluding Class F (with June average), we have in 

 A and ( ' the instances proving the rule ; while B, I> and F 

 contain the exceptions. 



Accordingly we find : in the first column, 33 instances, 

 7 exceptions ; in the second column, 89 instances, .3 ex- 

 ceptions. Further, it appears that, in 12 years out of 19 

 (or nearly 2 out of 3), with June wet, the rest of the 

 season was wet ; and in 16 years out of 21 (or about 4 out 

 of 5), with June dry, the rest of the season was dry. (The 

 corresponding figures for all days of rain are 13 out of 21, 

 and 10 out of 23.) 



It will be asked. Do these facts at all help us, at the end 

 of June, to forecast more definitely the character (number 

 of wet days) of the next three months ? I think they do. 



If we group together all the Junes which had the same 

 number of wet days, and compare, for each group, the 



