抄録
Close inspection reveals certain irregularities in daily milk yields in a cow. ll-)(jdia?ram uiven in Fi?. 1(I) shows a fairlv irregular cJnan?e in lactation, while it exhibitsa eeneral tendencv to show a smooth, flowin? increase and/or decrease. Stactsttcattreatment seems to be a best method to disentanule such at>rarentlv complicated seriesof varieties. Time serial analvses of successive milk vields were conductect on true aatasiven in Fi?. 1(1) in the followinu manner: An arithmetic movin? averawe of 31 days wascalculated for the oriuinal series of vields to obtain the r>lots of Fi?. I(2). The plotsof movin? average revealed a ratcid increase after rcarturition, a rise to its highest pointin a few months later, and a ?radual decrease to its termination as the calf grew. Theseries of riots in Flu. l(3) rerresents the remainder obtainccd bv subtracting the series ortclots in Fi?. l(2) from the oriuinal series in Flu. l(l). Thus, the oriuinal series of milkyxelds XVIS dlV1dC)d 11l1t0 tXV0 sernes; 011(If serues presennts the undulatory Ch2LIlg durtng awhole lactation time, and the other the temtcorar>or stcontaneous change. The tatteris a chaotic temrorarv chan?e and mav be divided further into two or three componentsof fluctuation>each of which has a rather different length of r>eriodicnty as aeterrntneaby the adequate mathematical procedure in the time serial analysis.If the isolated components are called the annual, monthly, weekly, and dailyfluctuatmOn 111 the order Of the respectmve PCIl0dlC tlllltt It IS P0SS1bIC [0 unravel suchcomplicated outside-world influences as climate agents, food intake, and bodily exerciseupon the rnilk secretion of a cow.The t>resent rarer is esreciall>concerned with analvtical consideration on tt-merelationship between the apparent spontaneous change of daily lactation and the air-temperature possibly functioning on it.l) The diasram of successive chances of dailv milk vields and averaue air-temperaturewas decomposed into the above-mentioned four types of fluctuations by 3) Influences of air-temperature i.e. heat condition upon the changes of lactationmay be determined by visual comparison of the phase between both components offluctuation having a similar periodic time, or otherwise by calculation of correlationcoefficients between the corresponding fluctuations of similar periodicity. It is a matterof course that the year-long change of both has been previously eliminated from theoriginal values.4) The above-mentioned technique of analysis was actually applied to the discus-sion in the present investigation. The subjects used were the daily milk yields offour Holstein cows kept on the Livestock Farm, University of Tokyo, in Ibaraki Prefectureand the records of daily average air-temperatures measured at the Meteorological Stationof Mito, 14 kilometers away from this farm. As a result, it was found that air-temperature above 20C in summer and below 5C in winter acted as a depressingagent on milk yield for the following one or two days. Furthermore, the day-to-daymilk yields joined alternatively with one another ; that is, any two adjacent milk yieldsmaintained a certain negative relationship in their absolute values.
