SATELLITE-TRACKING PROGRAIM — ^HAYES 353 



one could have realized all the intricacies of atmospheric density deter- 

 mined by seasonal, diurnal, and solar variations. Indeed, had it been 

 possible to predict these intricacies, there would have been consider- 

 ably less purpose to any passive tracking system. 



In the months that followed the laimchmg of Sputnik I the compu- 

 tations staff of the Observatory had a very difficult and very busy 

 time. They rapidly gained the experience that they all had lacked, 

 experience that had been impossible before a satellite was actually 

 launched. They realized early the magnitude of the job before them, 

 and were particularly conscious of the importance of air drag in the 

 computing of satellite predictions. All of them were mider constant 

 and heavy pressure, not only to organize an efficient means of gener- 

 ating predictions but also to help establish and maintain good rela- 

 tions with a now somewhat doubting public. Most of them during 

 those first weeks worked as much as 18 hours a day. Cots were set 

 up at Kittredge Hall, so that many of the computers simply did 

 not go home at all until they had achieved a basic and necessary 

 success. 



Their diligence and devotion were matched by those of other mem- 

 bers of the Observatory staff, and especially of the wives of all of 

 them. Headed by Mrs. Whipple and Mrs. Hynek, these good women 

 maintained a constant supply of coffee, sandwiches, clean shirts, and 

 other necessities. 



The first observations were, to say the least, rather inconsistent; 

 that is, the format and the data were not the same from one to the 

 next. In addition, some of these observations came from places that 

 had not been adequately "located" ; for example, if someone informed 

 the Observatory that he had witnessed a transit of the satellite, the 

 computers had to fuid out as exactly as possible the coordinates of 

 liis position. There was, then, a complex job of the bookkeeping, 

 as well as an equally complex task of reducing the data to a consistent 

 format. 



In addition to its own scientists and technicians the Observatory 

 called on mathematicians and astronomers of the Harvard staff, par- 

 ticularly Drs. Frances Wright and Eichard McCrosky, to help during 

 these first stages. 



Jack Slowey, Eobert Briggs, and Dr. John Eossoni of IBM soon 

 had the initial orbit program in operation. Through the traditional 

 Harvard Aimouncement Card (No. 1375) a preliminary estimate of 

 the orbit of Sputnik I was published on October 15. 



In theory, an orbit can be predicted from a set of any three observa- 

 tions. In their urgency to derive the orbit of Sputnik I, mathemati- 

 cians of the Observatory took such a set of three observations and 

 fed it through the initial orbit program. When the results did not 

 seem to match their estimate of the orbit, they rejected it and tried 



