1926: September 29, 1926

At 9 a.m. on September 29 Helen began her work at Harvard. She had a meeting with Shapley and she was going to help him with a book on star clusters.

In her very first weeks at the observatory, Shapley instructed her to get out various series of plates to estimate integrated diameters and magnitudes of globular clusters.

(Reported by Helen Sawyer Hogg in a talk, Shapley's Era, published in 1986, IAU Symposium, 126, 11).

1927: May 1, 1927

The Galactic Cluster NGC 6231 - a paper published by Shapley and Sawyer in 1927, Harv. Bulletin, 846,1
[This was the first of a series of papers that Helen wrote with Shapley. I don't know the precise date when it was submitted. Perhaps one can assume it was a month or two before publication.]

For this paper, Helen estimated magnitudes for 190 stars.

It appears that Shapley assigned this project to see how well she could estimate stellar magnitudes before they published the important work of determining photographic magnitudes of globular clusters and their stars.

1927: June 1, 1927

Note on the Supposed Variable Star TX Scorpii - a paper published by Helen in 1927, Harv. Bulletin, 847, 10.

Someone had made 63 visual observations of this star, derived a period ~0.94 days, suggesting the star was a Cepheid and published the results in the AJ.
This would have been an unusual object so Helen assessed its magnitude on 200 Harvard photographs made during the years 1899 to 1921 and found no light variation.

1927: July 1, 1927

Photographic magnitudes of 95 Globular Clusters - a paper published by Sawyer and Shapley in 1927, Harv. Bulletin, 848, 1.

In Shapley's earlier work at Mt. Wilson (1920, Mt. W. Contr. 190), he had determined that the mean absolute visual magnitude of a GC was -8.8 with an average deviation of half a mag.
Thus, for clusters to which this mean value applies, a determination of the apparent magnitude provides an easy method for determining the distance.

In this investigation, they published apparent photographic (integrated) magnitudes for 95 globular clusters estimated on Harvard photographs with reference to various Harvard magnitude sequences, based on the present international standards of photographic magnitude. The scale of the plates was 1mm = 10 arcminutes so that the great majority of the GC images could be easily compared with nearby star images.
Each cluster was measured on at least two plates and each plate was twice independently measured. An attempt was made to make the whole series of measures homogeneous. However, this was difficult for clusters brighter than 6th or 7th magnitude because they do not appear as star-like point sources.

They did not draw any conclusions. They stated that their integrated magnitudes would be discussed later, with respect to other properties of clusters.

1927: August 1, 1927

A Classification of Globular Clusters - a paper published by Shapley and Sawyer in 1927, Harv. Bulletin, 849, 11

This classification scheme was designed as a measure of the central concentration of a cluster on a scale of I (most centrally concentrated) to XII (loosest). There were 95 clusters in their sample and they each classified each cluster twice before adopting the final value.
In their paper, they discussed some of the difficulties, e.g. when a star is superimposed on the field.
They found that the various classes were widely spread in apparent brightness and diameter and did not depend on the integrated magnitude, except for a slight tendency of the least condensed clusters in the mean to be faint.
They postulated that the classes might be an indication of developmental age.

1927: September 6-8, 1927

The 38th AAS meeting was held in Madison, Wisconsin (1931, PAAS, 6, 19) and Helen (Sawyer) and Frank Hogg (her future husband) were both elected to membership.

Their addresses were listed as Harvard College Observatory, Cambridge, Mass. The total AAS membership was then 443. As far as I can tell, they did not attend the meeting. Neither of them presented a paper.

1927: November 1, 1927

Apparent Diameters and Ellipticities of Globular Clusters - paper published by Shapley and Sawyer in 1927, Harv. Bulletin, 852, 22

They pointed out that angular diameters of clusters are useful for determining relative distances.
Previous determinations of angular diameters had been derived mainly from paper or glass prints of Franklin Adams plates (a photographic atlas of the sky). It was better to use homogeneous series of direct photographs. This new study was based on three series of Harvard plates.
They made a plot of angular diameter vs. integrated magnitude and it showed a reasonably good correlation, but there was a systematic deviation and conspicuous scattering for clusters of concentration classes IX to XII.
Ellipticities were also determined and they noted their measures were probably much better than any previously made, but not of much significance except in the case of the few clusters that were the most elliptical.

They noted that in a forthcoming paper, they planned to derive revised distances, using the material from this paper, along with the available data on variable stars and stars of high luminosity.

1927: December 29-30, 1927

The 39th AAS meeting was held at Yale in New Haven, Conn (1931, PAAS, 6, 95)
Helen attended the meeting but did not present a paper.

However, discussions at that meeting inspired her to start her life-long research on variable stars in globular clusters.

At the Yale meeting Jan Schilt gave a paper on "The distribution of light in the globular cluster Omega Centauri". In the discussion that followed, he stated that he was tired of hearing astronomers talk about the period-luminosity relation in globular clusters. He declared firmly that it was based on so few clusters and so few variables that it had little meaning.

When Helen returned to Harvard after the meeting she started to collect all the references to globular clusters and the variables in them, something she continued to do all her life. She discovered that Schilt had been right so she embarked on a course of determining periods in little-studied clusters.

Up until 1932, she worked on Harvard plates. However, after her move to Victoria, BC in late August 1931, she started her own observing program at the DAO. She later continued it at the DDO when she moved to Richmond Hill, Ontario in 1935.

Helen related the story about Schilt in her outgoing presidential address to the Canadian Astronomical Society on May 12, 1972 (1973, JRASC, 67, 8).

[Note by CC: It appears that Schilt was referring to a paper by Shapley (1918, ApJ, 48, p89) in which he published a period-luminosity relation that combined data for SMC and globular cluster Cepheids. However, the only cluster that had (long period) Cepheids with a range of periods was Omega Centauri which had 5, with periods from approximately 1.5 to 29 days.

After Helen set up her own observing program, she discovered long period Cepheids in numerous other clusters. Her first major paper based on her own observations data was a study of M2 (1935, DAO Publications., Vol. 6, No. 14). She discovered 4 long period Cepheids in the cluster and plotted a a period-luminosity relation which fit the P-L relation that Shapley had derived earlier.]

1929: August 1 1929

The Distances of Ninety-three Globular Star Clusters - a paper published by Shapley and Sawyer in 1929, Harv. Bulletin, 869, 1

The investigation of variable stars in globular clusters was considered to be the fundamental method for deriving cluster distances. It was assumed that the mean absolute magnitude of the cluster variables (now known as RR Lyrae variables) was the same in all clusters. Thus, if the mean median magnitude of the RR Lyrae variables in a cluster could be established, the cluster's distance could be derived. Unfortunately, not all clusters have variable stars.
In Shapley's 1917 study at Mount Wilson (Mt. W. Contr. #152 = 1918, ApJ, 48, p154), only 5 of the 68 clusters he studied had variables with suitable data. In this paper, the number of clusters with variables had increased to 19. In these clusters, 594 variables had been studied enough to establish the mean median magnitudes of their cluster type variables. The distances derived from the variable stars were then used to calibrate the other methods of distance determination.

The second technique Shapley used for cluster distance determination in his earlier investigation was to measure the magnitudes of the brightest stars. In 1917, he had data for the brightest stars in 28 globular clusters. Now they had measures of the brightest stars in 48 clusters. The procedure they followed was to compute the difference between the median magnitude of the cluster variables and the mean magnitude of the brightest stars in each cluster that had variables. This was found to be relatively constant.

However, there were small variations that were related to the degree of concentration of the stars in a particular cluster. Because of this, for the analysis of the Harvard data (1927, Harvard Bulletin 849), Shapley and Sawyer introduced a classification scheme on a scale of I (most concentrated) to XII (loosest). Consequently, a cluster's concentration class could be taken into account when deriving a distance, based on the brightest stars.

In their earlier papers (Harvard Bull. 848 and 852), Shapley & Sawyer measured integrated magnitudes and angular diameters for more than 90 clusters. Both these properties are functions of the cluster distance.

Thus the data for the 48 clusters for which they derived distances based on variables and bright stars could be used to calibrate distances based on integrated magnitudes and angular diameters and made it possible to derive distances to 93 clusters.

When they finished deriving the distances, they compared the results with Shapley's (1917) Mount Wilson paper. They found the average difference was 12%, after correcting for a systematic decrease of 11%, which was due to a change in the zero point of the period-luminosity curve.

Because of the increase in basic photometric data and in the number of clusters with variables, they believed that the new values were much more secure than those formerly determined.

1929: December 1, 1929

Photometric Observations of TU Cassiopeia - paper published by Sawyer in 1929, Harv. Bulletin 871, 9

Helen made 174 photometric observations of this 2.139 day period Cepheid variable from November 22, 1928 to February 12, 1929. She confirmed previous observers' result that the star exhibits a double maximum.

1929: December 30, 1929 - January 2, 1930

The 43rd AAS meeting was held in Cambridge, Mass. Helen attended and presented a paper with Shapley: VARIABLE STARS IN GLOBULAR CLUSTERS.
The Abstract was published in 1930, Popular Ast., 38, 408 and again in 1931, Pub AAS, 6, 346

It summarized recent results. Of 93 GCs, 42 had been rather thoroughly examined for variables, 970 found and periods determined for 466 and among these, 355 were in only 4 clusters. Most of the variables appear to be cluster-type Cepheids (i.e., RR Lyrae type). The number of variables in a cluster had no known correlation with other data for the cluster, except for a slight tendency for the number of variables to decrease with approach to the galactic plane.
Mean median mag of cluster type variables is usually 1.2 mag fainter than the mean of the 25 brightest stars, as determined for 19 clusters. However, there were some peculiar clusters in which the variables were only 0.5 mag fainter.

1930: April 1, 1930

The Star Cluster NGC 2477 - paper published by Sawyer in 1930, Harv. Bulletin 875, 16

This was a southern Milky Way cluster that was unusual in symmetry and in number of stars.

Helen set up a magnitude sequence from 10.1 to 16.8 mag and estimated magnitudes of the cluster stars on plate obtained with the Bruce 24-inch. She plotted curves showing the number of stars in different magnitude ranges and concluded that the cluster might be an intermediate type between globular and galactic clusters. She searched for variables and found none. However, she commented that the existing plates were not suitable for the purpose so the search was inconclusive.

1930: June

Shapley finished his book Star Clusters.

Much of the material on globular clusters and their variable stars was based on work with Helen that had been published earlier in papers in the Harvard Bulletin, as noted above. Consequently, authors who cited their work usually cited Shapley's book and not the original papers.

In addition, Table IV, I, Summary of Variables in Clusters, on page 45, of Shapley's book was based on Helen's bibliographic efforts of several preceding years.
(She pointed this out in her presidential address to the Canadian Astronomical Society - 1973, JRASC, 67, 8.)

1930: September

Helen married Frank Hogg.

1931: May 1st

Helen submitted the paper, Periods and Light Curves of 32 Variable Stars in the Globular Clusters NGC 362, 6121 and 6397, published in 1931, Harvard Circular, 366

NGC 362: She studied the 14 variables discovered by Bailey and derived periods for 10 of them. To address the problem of which belonged to the SMC and which to NGC 362, she had discovered more variables between the two bodies, estimated magnitudes and planned to derive their periods and report the results in a later paper which she presented at the 47th AAS meeting in December 1931.

NGC 6121: She studied the 33 variables discovered by Leavitt and derived periods for 20.

NGC 6397: She derived periods for the two variables discovered by Bailey. Both had long periods and their membership status was uncertain.

Omega Centauri: She used additional observations to investigate the 5 long period Cepheids that had been studied earlier by Bailey and compared her results with his. She confirmed his periods for four of them: V1, V29, V48, V60, but was not able to derive a period for V61.

She also published a table listing all of the globular clusters containing cluster type variables of known period and discussed the period frequency distribution. She commented on the gap in period between Bailey's subclasses a, b, and c and commented that the gaps may prove of significance in cluster type variables.

In addition, she summarized existing data on variable stars with periods greater than one day in globular clusters and found that on a bolometric period-luminosity relation the long period variables continue the curve for [cluster] Cepheids, i.e. RR Lyrae variables.

1932: February 1st, 1932

Note on Luyten's Nova in the Large Magellanic Cloud - paper published by Sawyer in 1932, Harv. Bulletin 886, 15

By this time, Helen was settled in Victoria.
However, this was based on a measurement she made on a Harvard plate. She estimated the magnitude of the nova on a plate that was taken 20 days before Luyten's earliest plate and found it was at its maximum recorded magnitude, 12.4.
Presumably she submitted this note before she left New England. On the other hand, she might have taken the plate with her when she went to the DAO.