This paper describes a computer program used to compute equivalent widths via an IBM 2250 attached to an IBM 360/75J.

The measurement of stellar equivalent widths is a long tedious procedure and the final results depend to a large extent on where the stellar continuum is drawn. Therefore, a method which permits the continuum to be redrawn and the equivalent width remeasured quickly is desirable. An interactive CRT display device is an ideal instrument for allowing an astronomer to adjust the continuum level and the other parameters that are involved in equivalent width measurements. The display device discussed in this paper is an IBM model 2250 used in conjunction with a IBM 360/75J.

The “2250” consists of a 12 inch (30 cm) square CRT display tube plus a 32 function key box, an alphanumeric keyboard and a light pen which can be pointed at any illuminated spot on the screen. Thus, there are 34 interrupts that can be sensed by the program controlling the “2250”. The program that has been used communicates with the astronomer via messages and questions displayed on the CRT screen. The main program remains in a wait state until one of the 34 possible interrupts occurs, whereupon it branches to the designated subroutine and executes the procedures necessary to satisfy the interrupt.

During the 15 months that we operated the Celescope experiment on the Orbiting Astronomical Observatory, we acquired 8700 television frames of star fields containing a total of approximately 10 000 stars.

Figure 1 shows an example of one of these frames. Note the target ring, an aluminum deposit on the target of the tube, in each corner of the picture; the shadow in the upper left corner; the calibration lamp just below the center of the picture; and its ghost slightly below and to the left. All other objects are stars. Each frame consists of 256 scan lines designated by the number k, with each line containing 251 pixels (picture elements) designated by the number l, making a total of 65 000 intensity points I(k, l). The frames are divided into two spectral regions by two filters—lines 1 to 128 have one spectral range, lines 129 to 256 have a different one. Hence, each frame is reduced as two half frames in the first portion of the reduction system.

At Lick Observatory we have begun to use small computers as a basic element of some electronic instrumentation systems. It is felt that computer technology has advanced to the point where the results are likely to repay the investment. In our approach and choice of machine, we have tried to benefit by the experience gained over the past several years by nuclear physicists in nuclear spectroscopy applications.

The major factor involved in choosing a small computer is probably one of economics. The initial cost is important, but, equally important, the costs of maintenance, programming, and the connection of peripheral equipment must be considered. A computer is sufficiently complex to require at least one person in an organization to spend considerable time familiarizing himself with the details of the machine. These extra costs, required to realize the potential of any computer system, are likely to exceed the initial purchase price of the system! It is important to take such factors into account when choosing a machine.

A special matching facilities grant obtained from the National Science Foundation, in connection with the construction of the 107-inch reflector at McDonald Observatory, included the cost of an IBM 1800 process control computer. This computer was delivered in December, 1967, and has been in routine operation at the Observatory for slightly more than one year. The original system plan was presented as a paper at the meeting of the A.S.P. in June, 1968 and is shown in symbolic form in Figure 1. The essential features of the design were that it be advisory in nature and operate on a time-shared basis to provide assistance to the 107-inch, 82-inch and 36-inch telescopes. Facilities for automatic data acquisition were deferred until this initial plan was completed.